Showing 1–50 of 85 results for author: Zhai, S

Can Large Language Models Understand DL-Lite Ontologies? An Empirical Study

Authors: Keyu Wang, Guilin Qi, Jiaqi Li, Songlin Zhai

Abstract: Large language models (LLMs) have shown significant achievements in solving a wide range of tasks. Recently, LLMs' capability to store, retrieve and infer with symbolic knowledge has drawn a great deal of attention, showing their potential to understand structured information. However, it is not yet known whether LLMs can understand Description Logic (DL) ontologies. In this work, we empirically a… ▽ More Large language models (LLMs) have shown significant achievements in solving a wide range of tasks. Recently, LLMs' capability to store, retrieve and infer with symbolic knowledge has drawn a great deal of attention, showing their potential to understand structured information. However, it is not yet known whether LLMs can understand Description Logic (DL) ontologies. In this work, we empirically analyze the LLMs' capability of understanding DL-Lite ontologies covering 6 representative tasks from syntactic and semantic aspects. With extensive experiments, we demonstrate both the effectiveness and limitations of LLMs in understanding DL-Lite ontologies. We find that LLMs can understand formal syntax and model-theoretic semantics of concepts and roles. However, LLMs struggle with understanding TBox NI transitivity and handling ontologies with large ABoxes. We hope that our experiments and analyses provide more insights into LLMs and inspire to build more faithful knowledge engineering solutions. △ Less

Submitted 25 June, 2024; originally announced June 2024.

PGSR: Planar-based Gaussian Splatting for Efficient and High-Fidelity Surface Reconstruction

Authors: Danpeng Chen, Hai Li, Weicai Ye, Yifan Wang, Weijian Xie, Shangjin Zhai, Nan Wang, Haomin Liu, Hujun Bao, Guofeng Zhang

Abstract: Recently, 3D Gaussian Splatting (3DGS) has attracted widespread attention due to its high-quality rendering, and ultra-fast training and rendering speed. However, due to the unstructured and irregular nature of Gaussian point clouds, it is difficult to guarantee geometric reconstruction accuracy and multi-view consistency simply by relying on image reconstruction loss. Although many studies on sur… ▽ More Recently, 3D Gaussian Splatting (3DGS) has attracted widespread attention due to its high-quality rendering, and ultra-fast training and rendering speed. However, due to the unstructured and irregular nature of Gaussian point clouds, it is difficult to guarantee geometric reconstruction accuracy and multi-view consistency simply by relying on image reconstruction loss. Although many studies on surface reconstruction based on 3DGS have emerged recently, the quality of their meshes is generally unsatisfactory. To address this problem, we propose a fast planar-based Gaussian splatting reconstruction representation (PGSR) to achieve high-fidelity surface reconstruction while ensuring high-quality rendering. Specifically, we first introduce an unbiased depth rendering method, which directly renders the distance from the camera origin to the Gaussian plane and the corresponding normal map based on the Gaussian distribution of the point cloud, and divides the two to obtain the unbiased depth. We then introduce single-view geometric, multi-view photometric, and geometric regularization to preserve global geometric accuracy. We also propose a camera exposure compensation model to cope with scenes with large illumination variations. Experiments on indoor and outdoor scenes show that our method achieves fast training and rendering while maintaining high-fidelity rendering and geometric reconstruction, outperforming 3DGS-based and NeRF-based methods. △ Less

Submitted 10 June, 2024; originally announced June 2024.

Comments: project page: https://zju3dv.github.io/pgsr/

Proofread: Fixes All Errors with One Tap

Authors: Renjie Liu, Yanxiang Zhang, Yun Zhu, Haicheng Sun, Yuanbo Zhang, Michael Xuelin Huang, Shanqing Cai, Lei Meng, Shumin Zhai

Abstract: The impressive capabilities in Large Language Models (LLMs) provide a powerful approach to reimagine users' typing experience. This paper demonstrates Proofread, a novel Gboard feature powered by a server-side LLM in Gboard, enabling seamless sentence-level and paragraph-level corrections with a single tap. We describe the complete system in this paper, from data generation, metrics design to mode… ▽ More The impressive capabilities in Large Language Models (LLMs) provide a powerful approach to reimagine users' typing experience. This paper demonstrates Proofread, a novel Gboard feature powered by a server-side LLM in Gboard, enabling seamless sentence-level and paragraph-level corrections with a single tap. We describe the complete system in this paper, from data generation, metrics design to model tuning and deployment. To obtain models with sufficient quality, we implement a careful data synthetic pipeline tailored to online use cases, design multifaceted metrics, employ a two-stage tuning approach to acquire the dedicated LLM for the feature: the Supervised Fine Tuning (SFT) for foundational quality, followed by the Reinforcement Learning (RL) tuning approach for targeted refinement. Specifically, we find sequential tuning on Rewrite and proofread tasks yields the best quality in SFT stage, and propose global and direct rewards in the RL tuning stage to seek further improvement. Extensive experiments on a human-labeled golden set showed our tuned PaLM2-XS model achieved 85.56\% good ratio. We launched the feature to Pixel 8 devices by serving the model on TPU v5 in Google Cloud, with thousands of daily active users. Serving latency was significantly reduced by quantization, bucket inference, text segmentation, and speculative decoding. Our demo could be seen in \href{https://youtu.be/4ZdcuiwFU7I}{Youtube}. △ Less

Submitted 6 June, 2024; originally announced June 2024.

Comments: 8 pages, 3 figures, 2 tables

Physics-Informed Neural Networks for Dynamic Process Operations with Limited Physical Knowledge and Data

Authors: Mehmet Velioglu, Song Zhai, Sophia Rupprecht, Alexander Mitsos, Andreas Jupke, Manuel Dahmen

Abstract: In chemical engineering, process data are expensive to acquire, and complex phenomena are difficult to fully model. We explore the use of physics-informed neural networks (PINNs) for dynamic processes with incomplete mechanistic semi-explicit differential-algebraic equation systems and scarce process data. In particular, we focus on estimating states for which neither direct observational data nor… ▽ More In chemical engineering, process data are expensive to acquire, and complex phenomena are difficult to fully model. We explore the use of physics-informed neural networks (PINNs) for dynamic processes with incomplete mechanistic semi-explicit differential-algebraic equation systems and scarce process data. In particular, we focus on estimating states for which neither direct observational data nor constitutive equations are available. We propose an easy-to-apply heuristic to assess whether estimation of such states may be possible. As numerical examples, we consider a continuously stirred tank reactor and a liquid-liquid separator. We find that PINNs can infer unmeasured states with reasonable accuracy, and they generalize better in low-data scenarios than purely data-driven models. We thus show that PINNs are capable of modeling processes when relatively few experimental data and only partially known mechanistic descriptions are available, and conclude that they constitute a promising avenue that warrants further investigation. △ Less

Submitted 7 July, 2024; v1 submitted 3 June, 2024; originally announced June 2024.

Comments: manuscript (32 pages, 9 figures, 11 tables), supporting materials (14 pages, 4 figures, 5 tables)

Improving GFlowNets for Text-to-Image Diffusion Alignment

Authors: Dinghuai Zhang, Yizhe Zhang, Jiatao Gu, Ruixiang Zhang, Josh Susskind, Navdeep Jaitly, Shuangfei Zhai

Abstract: Diffusion models have become the de-facto approach for generating visual data, which are trained to match the distribution of the training dataset. In addition, we also want to control generation to fulfill desired properties such as alignment to a text description, which can be specified with a black-box reward function. Prior works fine-tune pretrained diffusion models to achieve this goal throu… ▽ More Diffusion models have become the de-facto approach for generating visual data, which are trained to match the distribution of the training dataset. In addition, we also want to control generation to fulfill desired properties such as alignment to a text description, which can be specified with a black-box reward function. Prior works fine-tune pretrained diffusion models to achieve this goal through reinforcement learning-based algorithms. Nonetheless, they suffer from issues including slow credit assignment as well as low quality in their generated samples. In this work, we explore techniques that do not directly maximize the reward but rather generate high-reward images with relatively high probability -- a natural scenario for the framework of generative flow networks (GFlowNets). To this end, we propose the Diffusion Alignment with GFlowNet (DAG) algorithm to post-train diffusion models with black-box property functions. Extensive experiments on Stable Diffusion and various reward specifications corroborate that our method could effectively align large-scale text-to-image diffusion models with given reward information. △ Less

Submitted 16 June, 2024; v1 submitted 2 June, 2024; originally announced June 2024.

Kaleido Diffusion: Improving Conditional Diffusion Models with Autoregressive Latent Modeling

Authors: Jiatao Gu, Ying Shen, Shuangfei Zhai, Yizhe Zhang, Navdeep Jaitly, Joshua M. Susskind

Abstract: Diffusion models have emerged as a powerful tool for generating high-quality images from textual descriptions. Despite their successes, these models often exhibit limited diversity in the sampled images, particularly when sampling with a high classifier-free guidance weight. To address this issue, we present Kaleido, a novel approach that enhances the diversity of samples by incorporating autoregr… ▽ More Diffusion models have emerged as a powerful tool for generating high-quality images from textual descriptions. Despite their successes, these models often exhibit limited diversity in the sampled images, particularly when sampling with a high classifier-free guidance weight. To address this issue, we present Kaleido, a novel approach that enhances the diversity of samples by incorporating autoregressive latent priors. Kaleido integrates an autoregressive language model that encodes the original caption and generates latent variables, serving as abstract and intermediary representations for guiding and facilitating the image generation process. In this paper, we explore a variety of discrete latent representations, including textual descriptions, detection bounding boxes, object blobs, and visual tokens. These representations diversify and enrich the input conditions to the diffusion models, enabling more diverse outputs. Our experimental results demonstrate that Kaleido effectively broadens the diversity of the generated image samples from a given textual description while maintaining high image quality. Furthermore, we show that Kaleido adheres closely to the guidance provided by the generated latent variables, demonstrating its capability to effectively control and direct the image generation process. △ Less

Submitted 31 May, 2024; originally announced May 2024.

Comments: 22 pages, 14 figures

Membership Inference on Text-to-Image Diffusion Models via Conditional Likelihood Discrepancy

Authors: Shengfang Zhai, Huanran Chen, Yinpeng Dong, Jiajun Li, Qingni Shen, Yansong Gao, Hang Su, Yang Liu

Abstract: Text-to-image diffusion models have achieved tremendous success in the field of controllable image generation, while also coming along with issues of privacy leakage and data copyrights. Membership inference arises in these contexts as a potential auditing method for detecting unauthorized data usage. While some efforts have been made on diffusion models, they are not applicable to text-to-image d… ▽ More Text-to-image diffusion models have achieved tremendous success in the field of controllable image generation, while also coming along with issues of privacy leakage and data copyrights. Membership inference arises in these contexts as a potential auditing method for detecting unauthorized data usage. While some efforts have been made on diffusion models, they are not applicable to text-to-image diffusion models due to the high computation overhead and enhanced generalization capabilities. In this paper, we first identify a conditional overfitting phenomenon in text-to-image diffusion models, indicating that these models tend to overfit the conditional distribution of images given the text rather than the marginal distribution of images. Based on this observation, we derive an analytical indicator, namely Conditional Likelihood Discrepancy (CLiD), to perform membership inference, which reduces the stochasticity in estimating the memorization of individual samples. Experimental results demonstrate that our method significantly outperforms previous methods across various data distributions and scales. Additionally, our method shows superior resistance to overfitting mitigation strategies such as early stopping and data augmentation. △ Less

Submitted 29 May, 2024; v1 submitted 23 May, 2024; originally announced May 2024.

Comments: 17 pages, 5 figures; minor typos corrected

Many-to-many Image Generation with Auto-regressive Diffusion Models

Authors: Ying Shen, Yizhe Zhang, Shuangfei Zhai, Lifu Huang, Joshua M. Susskind, Jiatao Gu

Abstract: Recent advancements in image generation have made significant progress, yet existing models present limitations in perceiving and generating an arbitrary number of interrelated images within a broad context. This limitation becomes increasingly critical as the demand for multi-image scenarios, such as multi-view images and visual narratives, grows with the expansion of multimedia platforms. This p… ▽ More Recent advancements in image generation have made significant progress, yet existing models present limitations in perceiving and generating an arbitrary number of interrelated images within a broad context. This limitation becomes increasingly critical as the demand for multi-image scenarios, such as multi-view images and visual narratives, grows with the expansion of multimedia platforms. This paper introduces a domain-general framework for many-to-many image generation, capable of producing interrelated image series from a given set of images, offering a scalable solution that obviates the need for task-specific solutions across different multi-image scenarios. To facilitate this, we present MIS, a novel large-scale multi-image dataset, containing 12M synthetic multi-image samples, each with 25 interconnected images. Utilizing Stable Diffusion with varied latent noises, our method produces a set of interconnected images from a single caption. Leveraging MIS, we learn M2M, an autoregressive model for many-to-many generation, where each image is modeled within a diffusion framework. Throughout training on the synthetic MIS, the model excels in capturing style and content from preceding images - synthetic or real - and generates novel images following the captured patterns. Furthermore, through task-specific fine-tuning, our model demonstrates its adaptability to various multi-image generation tasks, including Novel View Synthesis and Visual Procedure Generation. △ Less

Submitted 3 April, 2024; originally announced April 2024.

How Far Are We from Intelligent Visual Deductive Reasoning?

Authors: Yizhe Zhang, He Bai, Ruixiang Zhang, Jiatao Gu, Shuangfei Zhai, Josh Susskind, Navdeep Jaitly

Abstract: Vision-Language Models (VLMs) such as GPT-4V have recently demonstrated incredible strides on diverse vision language tasks. We dig into vision-based deductive reasoning, a more sophisticated but less explored realm, and find previously unexposed blindspots in the current SOTA VLMs. Specifically, we leverage Raven's Progressive Matrices (RPMs), to assess VLMs' abilities to perform multi-hop relati… ▽ More Vision-Language Models (VLMs) such as GPT-4V have recently demonstrated incredible strides on diverse vision language tasks. We dig into vision-based deductive reasoning, a more sophisticated but less explored realm, and find previously unexposed blindspots in the current SOTA VLMs. Specifically, we leverage Raven's Progressive Matrices (RPMs), to assess VLMs' abilities to perform multi-hop relational and deductive reasoning relying solely on visual clues. We perform comprehensive evaluations of several popular VLMs employing standard strategies such as in-context learning, self-consistency, and Chain-of-thoughts (CoT) on three diverse datasets, including the Mensa IQ test, IntelligenceTest, and RAVEN. The results reveal that despite the impressive capabilities of LLMs in text-based reasoning, we are still far from achieving comparable proficiency in visual deductive reasoning. We found that certain standard strategies that are effective when applied to LLMs do not seamlessly translate to the challenges presented by visual reasoning tasks. Moreover, a detailed analysis reveals that VLMs struggle to solve these tasks mainly because they are unable to perceive and comprehend multiple, confounding abstract patterns in RPM examples. △ Less

Submitted 8 March, 2024; v1 submitted 7 March, 2024; originally announced March 2024.

Comments: ICLR 2024 AGI workshop. https://github.com/apple/ml-rpm-bench

Discovering Universal Semantic Triggers for Text-to-Image Synthesis

Authors: Shengfang Zhai, Weilong Wang, Jiajun Li, Yinpeng Dong, Hang Su, Qingni Shen

Abstract: Recently text-to-image models have gained widespread attention in the community due to their controllable and high-quality generation ability. However, the robustness of such models and their potential ethical issues have not been fully explored. In this paper, we introduce Universal Semantic Trigger, a meaningless token sequence that can be added at any location within the input text yet can indu… ▽ More Recently text-to-image models have gained widespread attention in the community due to their controllable and high-quality generation ability. However, the robustness of such models and their potential ethical issues have not been fully explored. In this paper, we introduce Universal Semantic Trigger, a meaningless token sequence that can be added at any location within the input text yet can induce generated images towards a preset semantic target.To thoroughly investigate it, we propose Semantic Gradient-based Search (SGS) framework. SGS automatically discovers the potential universal semantic triggers based on the given semantic targets. Furthermore, we design evaluation metrics to comprehensively evaluate semantic shift of images caused by these triggers. And our empirical analyses reveal that the mainstream open-source text-to-image models are vulnerable to our triggers, which could pose significant ethical threats. Our work contributes to a further understanding of text-to-image synthesis and helps users to automatically auditing their models before deployment. △ Less

Submitted 12 February, 2024; originally announced February 2024.

Comments: 9 pages, 5 figures. Work in progress

Rambler: Supporting Writing With Speech via LLM-Assisted Gist Manipulation

Authors: Susan Lin, Jeremy Warner, J. D. Zamfirescu-Pereira, Matthew G. Lee, Sauhard Jain, Michael Xuelin Huang, Piyawat Lertvittayakumjorn, Shanqing Cai, Shumin Zhai, Björn Hartmann, Can Liu

Abstract: Dictation enables efficient text input on mobile devices. However, writing with speech can produce disfluent, wordy, and incoherent text and thus requires heavy post-processing. This paper presents Rambler, an LLM-powered graphical user interface that supports gist-level manipulation of dictated text with two main sets of functions: gist extraction and macro revision. Gist extraction generates key… ▽ More Dictation enables efficient text input on mobile devices. However, writing with speech can produce disfluent, wordy, and incoherent text and thus requires heavy post-processing. This paper presents Rambler, an LLM-powered graphical user interface that supports gist-level manipulation of dictated text with two main sets of functions: gist extraction and macro revision. Gist extraction generates keywords and summaries as anchors to support the review and interaction with spoken text. LLM-assisted macro revisions allow users to respeak, split, merge and transform dictated text without specifying precise editing locations. Together they pave the way for interactive dictation and revision that help close gaps between spontaneous spoken words and well-structured writing. In a comparative study with 12 participants performing verbal composition tasks, Rambler outperformed the baseline of a speech-to-text editor + ChatGPT, as it better facilitates iterative revisions with enhanced user control over the content while supporting surprisingly diverse user strategies. △ Less

Submitted 7 March, 2024; v1 submitted 19 January, 2024; originally announced January 2024.

Comments: To appear at ACM CHI 2024

Scalable Pre-training of Large Autoregressive Image Models

Authors: Alaaeldin El-Nouby, Michal Klein, Shuangfei Zhai, Miguel Angel Bautista, Alexander Toshev, Vaishaal Shankar, Joshua M Susskind, Armand Joulin

Abstract: This paper introduces AIM, a collection of vision models pre-trained with an autoregressive objective. These models are inspired by their textual counterparts, i.e., Large Language Models (LLMs), and exhibit similar scaling properties. Specifically, we highlight two key findings: (1) the performance of the visual features scale with both the model capacity and the quantity of data, (2) the value o… ▽ More This paper introduces AIM, a collection of vision models pre-trained with an autoregressive objective. These models are inspired by their textual counterparts, i.e., Large Language Models (LLMs), and exhibit similar scaling properties. Specifically, we highlight two key findings: (1) the performance of the visual features scale with both the model capacity and the quantity of data, (2) the value of the objective function correlates with the performance of the model on downstream tasks. We illustrate the practical implication of these findings by pre-training a 7 billion parameter AIM on 2 billion images, that achieves 84.0% on ImageNet-1k with a frozen trunk. Interestingly, even at this scale, we observe no sign of saturation in performance, suggesting that AIM potentially represents a new frontier for training large-scale vision models. The pre-training of AIM is similar to the pre-training of LLMs, and does not require any image-specific strategy to stabilize the training at scale. △ Less

Submitted 16 January, 2024; originally announced January 2024.

Comments: https://github.com/apple/ml-aim

TRLS: A Time Series Representation Learning Framework via Spectrogram for Medical Signal Processing

Authors: Luyuan Xie, Cong Li, Xin Zhang, Shengfang Zhai, Yuejian Fang, Qingni Shen, Zhonghai Wu

Abstract: Representation learning frameworks in unlabeled time series have been proposed for medical signal processing. Despite the numerous excellent progresses have been made in previous works, we observe the representation extracted for the time series still does not generalize well. In this paper, we present a Time series (medical signal) Representation Learning framework via Spectrogram (TRLS) to get m… ▽ More Representation learning frameworks in unlabeled time series have been proposed for medical signal processing. Despite the numerous excellent progresses have been made in previous works, we observe the representation extracted for the time series still does not generalize well. In this paper, we present a Time series (medical signal) Representation Learning framework via Spectrogram (TRLS) to get more informative representations. We transform the input time-domain medical signals into spectrograms and design a time-frequency encoder named Time Frequency RNN (TFRNN) to capture more robust multi-scale representations from the augmented spectrograms. Our TRLS takes spectrogram as input with two types of different data augmentations and maximizes the similarity between positive ones, which effectively circumvents the problem of designing negative samples. Our evaluation of four real-world medical signal datasets focusing on medical signal classification shows that TRLS is superior to the existing frameworks. △ Less

Submitted 5 January, 2024; originally announced January 2024.

Comments: This paper is accept by ICASSP 2024. This is a more detailed version

Building Open-Ended Embodied Agent via Language-Policy Bidirectional Adaptation

Authors: Shaopeng Zhai, Jie Wang, Tianyi Zhang, Fuxian Huang, Qi Zhang, Ming Zhou, Jing Hou, Yu Qiao, Yu Liu

Abstract: Building embodied agents on integrating Large Language Models (LLMs) and Reinforcement Learning (RL) have revolutionized human-AI interaction: researchers can now leverage language instructions to plan decision-making for open-ended tasks. However, existing research faces challenges in meeting the requirement of open-endedness. They typically either train LLM/RL models to adapt to a fixed counterp… ▽ More Building embodied agents on integrating Large Language Models (LLMs) and Reinforcement Learning (RL) have revolutionized human-AI interaction: researchers can now leverage language instructions to plan decision-making for open-ended tasks. However, existing research faces challenges in meeting the requirement of open-endedness. They typically either train LLM/RL models to adapt to a fixed counterpart, limiting exploration of novel skills and hindering the efficacy of human-AI interaction. To this end, we present OpenPAL, a co-training framework comprising two stages: (1) fine-tuning a pre-trained LLM to translate human instructions into goals for planning, and goal-conditioned training a policy for decision-making; (2) co-training to align the LLM and policy, achieving instruction open-endedness. We conducted experiments using Contra, an open-ended FPS game, demonstrating that an agent trained with OpenPAL not only comprehends arbitrary instructions but also exhibits efficient execution. These results suggest that OpenPAL holds the potential to construct open-ended embodied agents in practical scenarios. △ Less

Submitted 6 February, 2024; v1 submitted 12 December, 2023; originally announced January 2024.

Extended p-median problems for balancing service efficiency and equality

Authors: Yunfeng Kong, Chenchen Lian, Guangli Zhang, Shiyan Zhai

Abstract: This article deals with the location problem for balancing the service efficiency and equality. In public service systems, some people may feel envy in case that they need longer travel distance to access services than others. The strength of the envy can be measured by comparing one's travel distance to service facility with a threshold distance. Using the total envy function, four extended p-med… ▽ More This article deals with the location problem for balancing the service efficiency and equality. In public service systems, some people may feel envy in case that they need longer travel distance to access services than others. The strength of the envy can be measured by comparing one's travel distance to service facility with a threshold distance. Using the total envy function, four extended p-median problems are proposed for trade-off between service efficiency and equality. Five analytical properties of the new problems are mathematically proven. The new problems were tested on three sets of well-designed instances. The experimentation shows that the equality measures, such as the standard deviation, the mean absolute deviation, and the Gini coefficient between travel distances, can be substantially improved by minimizing the travel cost and the spatial envy. The experimentation also shows that, when the service supply is given in terms of the number of facilities, the service equality can be considerably improved by slightly increasing the travel distance. When the service supply is increased in terms of the number of facilities, both the service efficiency and spatial equality can be significantly improved. △ Less

Submitted 25 April, 2024; v1 submitted 21 December, 2023; originally announced December 2023.

Comments: 38 pages, 4 tables, 5 figures

MSC Class: 90C27 ACM Class: J.6

Mental Health Diagnosis in the Digital Age: Harnessing Sentiment Analysis on Social Media Platforms upon Ultra-Sparse Feature Content

Authors: Haijian Shao, Ming Zhu, Shengjie Zhai

Abstract: Amid growing global mental health concerns, particularly among vulnerable groups, natural language processing offers a tremendous potential for early detection and intervention of people's mental disorders via analyzing their postings and discussions on social media platforms. However, ultra-sparse training data, often due to vast vocabularies and low-frequency words, hinders the analysis accuracy… ▽ More Amid growing global mental health concerns, particularly among vulnerable groups, natural language processing offers a tremendous potential for early detection and intervention of people's mental disorders via analyzing their postings and discussions on social media platforms. However, ultra-sparse training data, often due to vast vocabularies and low-frequency words, hinders the analysis accuracy. Multi-labeling and Co-occurrences of symptoms may also blur the boundaries in distinguishing similar/co-related disorders. To address these issues, we propose a novel semantic feature preprocessing technique with a three-folded structure: 1) mitigating the feature sparsity with a weak classifier, 2) adaptive feature dimension with modulus loops, and 3) deep-mining and extending features among the contexts. With enhanced semantic features, we train a machine learning model to predict and classify mental disorders. We utilize the Reddit Mental Health Dataset 2022 to examine conditions such as Anxiety, Borderline Personality Disorder (BPD), and Bipolar-Disorder (BD) and present solutions to the data sparsity challenge, highlighted by 99.81% non-zero elements. After applying our preprocessing technique, the feature sparsity decreases to 85.4%. Overall, our methods, when compared to seven benchmark models, demonstrate significant performance improvements: 8.0% in accuracy, 0.069 in precision, 0.093 in recall, 0.102 in F1 score, and 0.059 in AUC. This research provides foundational insights for mental health prediction and monitoring, providing innovative solutions to navigate challenges associated with ultra-sparse data feature and intricate multi-label classification in the domain of mental health analysis. △ Less

Submitted 8 November, 2023; originally announced November 2023.

Matryoshka Diffusion Models

Authors: Jiatao Gu, Shuangfei Zhai, Yizhe Zhang, Josh Susskind, Navdeep Jaitly

Abstract: Diffusion models are the de facto approach for generating high-quality images and videos, but learning high-dimensional models remains a formidable task due to computational and optimization challenges. Existing methods often resort to training cascaded models in pixel space or using a downsampled latent space of a separately trained auto-encoder. In this paper, we introduce Matryoshka Diffusion M… ▽ More Diffusion models are the de facto approach for generating high-quality images and videos, but learning high-dimensional models remains a formidable task due to computational and optimization challenges. Existing methods often resort to training cascaded models in pixel space or using a downsampled latent space of a separately trained auto-encoder. In this paper, we introduce Matryoshka Diffusion Models(MDM), an end-to-end framework for high-resolution image and video synthesis. We propose a diffusion process that denoises inputs at multiple resolutions jointly and uses a NestedUNet architecture where features and parameters for small-scale inputs are nested within those of large scales. In addition, MDM enables a progressive training schedule from lower to higher resolutions, which leads to significant improvements in optimization for high-resolution generation. We demonstrate the effectiveness of our approach on various benchmarks, including class-conditioned image generation, high-resolution text-to-image, and text-to-video applications. Remarkably, we can train a single pixel-space model at resolutions of up to 1024x1024 pixels, demonstrating strong zero-shot generalization using the CC12M dataset, which contains only 12 million images. △ Less

Submitted 23 October, 2023; originally announced October 2023.

Comments: 28 pages, 18 figures

Generative Modeling with Phase Stochastic Bridges

Authors: Tianrong Chen, Jiatao Gu, Laurent Dinh, Evangelos A. Theodorou, Joshua Susskind, Shuangfei Zhai

Abstract: Diffusion models (DMs) represent state-of-the-art generative models for continuous inputs. DMs work by constructing a Stochastic Differential Equation (SDE) in the input space (ie, position space), and using a neural network to reverse it. In this work, we introduce a novel generative modeling framework grounded in \textbf{phase space dynamics}, where a phase space is defined as {an augmented spac… ▽ More Diffusion models (DMs) represent state-of-the-art generative models for continuous inputs. DMs work by constructing a Stochastic Differential Equation (SDE) in the input space (ie, position space), and using a neural network to reverse it. In this work, we introduce a novel generative modeling framework grounded in \textbf{phase space dynamics}, where a phase space is defined as {an augmented space encompassing both position and velocity.} Leveraging insights from Stochastic Optimal Control, we construct a path measure in the phase space that enables efficient sampling. {In contrast to DMs, our framework demonstrates the capability to generate realistic data points at an early stage of dynamics propagation.} This early prediction sets the stage for efficient data generation by leveraging additional velocity information along the trajectory. On standard image generation benchmarks, our model yields favorable performance over baselines in the regime of small Number of Function Evaluations (NFEs). Furthermore, our approach rivals the performance of diffusion models equipped with efficient sampling techniques, underscoring its potential as a new tool generative modeling. △ Less

Submitted 12 May, 2024; v1 submitted 11 October, 2023; originally announced October 2023.

GAME: Generalized deep learning model towards multimodal data integration for early screening of adolescent mental disorders

Authors: Zhicheng Du, Chenyao Jiang, Xi Yuan, Shiyao Zhai, Zhengyang Lei, Shuyue Ma, Yang Liu, Qihui Ye, Chufan Xiao, Qiming Huang, Ming Xu, Dongmei Yu, Peiwu Qin

Abstract: The timely identification of mental disorders in adolescents is a global public health challenge.Single factor is difficult to detect the abnormality due to its complex and subtle nature. Additionally, the generalized multimodal Computer-Aided Screening (CAS) systems with interactive robots for adolescent mental disorders are not available. Here, we design an android application with mini-games an… ▽ More The timely identification of mental disorders in adolescents is a global public health challenge.Single factor is difficult to detect the abnormality due to its complex and subtle nature. Additionally, the generalized multimodal Computer-Aided Screening (CAS) systems with interactive robots for adolescent mental disorders are not available. Here, we design an android application with mini-games and chat recording deployed in a portable robot to screen 3,783 middle school students and construct the multimodal screening dataset, including facial images, physiological signs, voice recordings, and textual transcripts.We develop a model called GAME (Generalized Model with Attention and Multimodal EmbraceNet) with novel attention mechanism that integrates cross-modal features into the model. GAME evaluates adolescent mental conditions with high accuracy (73.34%-92.77%) and F1-Score (71.32%-91.06%).We find each modality contributes dynamically to the mental disorders screening and comorbidities among various mental disorders, indicating the feasibility of explainable model. This study provides a system capable of acquiring multimodal information and constructs a generalized multimodal integration algorithm with novel attention mechanisms for the early screening of adolescent mental disorders. △ Less

Submitted 18 September, 2023; originally announced September 2023.

Depth Completion with Multiple Balanced Bases and Confidence for Dense Monocular SLAM

Authors: Weijian Xie, Guanyi Chu, Quanhao Qian, Yihao Yu, Hai Li, Danpeng Chen, Shangjin Zhai, Nan Wang, Hujun Bao, Guofeng Zhang

Abstract: Dense SLAM based on monocular cameras does indeed have immense application value in the field of AR/VR, especially when it is performed on a mobile device. In this paper, we propose a novel method that integrates a light-weight depth completion network into a sparse SLAM system using a multi-basis depth representation, so that dense mapping can be performed online even on a mobile phone. Specifica… ▽ More Dense SLAM based on monocular cameras does indeed have immense application value in the field of AR/VR, especially when it is performed on a mobile device. In this paper, we propose a novel method that integrates a light-weight depth completion network into a sparse SLAM system using a multi-basis depth representation, so that dense mapping can be performed online even on a mobile phone. Specifically, we present a specifically optimized multi-basis depth completion network, called BBC-Net, tailored to the characteristics of traditional sparse SLAM systems. BBC-Net can predict multiple balanced bases and a confidence map from a monocular image with sparse points generated by off-the-shelf keypoint-based SLAM systems. The final depth is a linear combination of predicted depth bases that can be optimized by tuning the corresponding weights. To seamlessly incorporate the weights into traditional SLAM optimization and ensure efficiency and robustness, we design a set of depth weight factors, which makes our network a versatile plug-in module, facilitating easy integration into various existing sparse SLAM systems and significantly enhancing global depth consistency through bundle adjustment. To verify the portability of our method, we integrate BBC-Net into two representative SLAM systems. The experimental results on various datasets show that the proposed method achieves better performance in monocular dense mapping than the state-of-the-art methods. We provide an online demo running on a mobile phone, which verifies the efficiency and mapping quality of the proposed method in real-world scenarios. △ Less

Submitted 20 September, 2023; v1 submitted 8 September, 2023; originally announced September 2023.

Prompt-enhanced Hierarchical Transformer Elevating Cardiopulmonary Resuscitation Instruction via Temporal Action Segmentation

Authors: Yang Liu, Xiaoyun Zhong, Shiyao Zhai, Zhicheng Du, Zhenyuan Gao, Qiming Huang, Canyang Zhang, Bin Jiang, Vijay Kumar Pandey, Sanyang Han, Runming Wang, Yuxing Han, Peiwu Qin

Abstract: The vast majority of people who suffer unexpected cardiac arrest are performed cardiopulmonary resuscitation (CPR) by passersby in a desperate attempt to restore life, but endeavors turn out to be fruitless on account of disqualification. Fortunately, many pieces of research manifest that disciplined training will help to elevate the success rate of resuscitation, which constantly desires a seamle… ▽ More The vast majority of people who suffer unexpected cardiac arrest are performed cardiopulmonary resuscitation (CPR) by passersby in a desperate attempt to restore life, but endeavors turn out to be fruitless on account of disqualification. Fortunately, many pieces of research manifest that disciplined training will help to elevate the success rate of resuscitation, which constantly desires a seamless combination of novel techniques to yield further advancement. To this end, we collect a custom CPR video dataset in which trainees make efforts to behave resuscitation on mannequins independently in adherence to approved guidelines, thereby devising an auxiliary toolbox to assist supervision and rectification of intermediate potential issues via modern deep learning methodologies. Our research empirically views this problem as a temporal action segmentation (TAS) task in computer vision, which aims to segment an untrimmed video at a frame-wise level. Here, we propose a Prompt-enhanced hierarchical Transformer (PhiTrans) that integrates three indispensable modules, including a textual prompt-based Video Features Extractor (VFE), a transformer-based Action Segmentation Executor (ASE), and a regression-based Prediction Refinement Calibrator (PRC). The backbone of the model preferentially derives from applications in three approved public datasets (GTEA, 50Salads, and Breakfast) collected for TAS tasks, which accounts for the excavation of the segmentation pipeline on the CPR dataset. In general, we unprecedentedly probe into a feasible pipeline that genuinely elevates the CPR instruction qualification via action segmentation in conjunction with cutting-edge deep learning techniques. Associated experiments advocate our implementation with multiple metrics surpassing 91.0%. △ Less

Submitted 31 August, 2023; originally announced August 2023.

Comments: Transformer for Cardiopulmonary Resuscitation

Object Detection for Caries or Pit and Fissure Sealing Requirement in Children's First Permanent Molars

Authors: Chenyao Jiang, Shiyao Zhai, Hengrui Song, Yuqing Ma, Yachen Fan, Yancheng Fang, Dongmei Yu, Canyang Zhang, Sanyang Han, Runming Wang, Yong Liu, Jianbo Li, Peiwu Qin

Abstract: Dental caries is one of the most common oral diseases that, if left untreated, can lead to a variety of oral problems. It mainly occurs inside the pits and fissures on the occlusal/buccal/palatal surfaces of molars and children are a high-risk group for pit and fissure caries in permanent molars. Pit and fissure sealing is one of the most effective methods that is widely used in prevention of pit… ▽ More Dental caries is one of the most common oral diseases that, if left untreated, can lead to a variety of oral problems. It mainly occurs inside the pits and fissures on the occlusal/buccal/palatal surfaces of molars and children are a high-risk group for pit and fissure caries in permanent molars. Pit and fissure sealing is one of the most effective methods that is widely used in prevention of pit and fissure caries. However, current detection of pits and fissures or caries depends primarily on the experienced dentists, which ordinary parents do not have, and children may miss the remedial treatment without timely detection. To address this issue, we present a method to autodetect caries and pit and fissure sealing requirements using oral photos taken by smartphones. We use the YOLOv5 and YOLOX models and adopt a tiling strategy to reduce information loss during image pre-processing. The best result for YOLOXs model with tiling strategy is 72.3 mAP.5, while the best result without tiling strategy is 71.2. YOLOv5s6 model with/without tiling attains 70.9/67.9 mAP.5, respectively. We deploy the pre-trained network to mobile devices as a WeChat applet, allowing in-home detection by parents or children guardian. △ Less

Submitted 31 August, 2023; originally announced August 2023.

Private Federated Learning in Gboard

Authors: Yuanbo Zhang, Daniel Ramage, Zheng Xu, Yanxiang Zhang, Shumin Zhai, Peter Kairouz

Abstract: This white paper describes recent advances in Gboard(Google Keyboard)'s use of federated learning, DP-Follow-the-Regularized-Leader (DP-FTRL) algorithm, and secure aggregation techniques to train machine learning (ML) models for suggestion, prediction and correction intelligence from many users' typing data. Gboard's investment in those privacy technologies allows users' typing data to be processe… ▽ More This white paper describes recent advances in Gboard(Google Keyboard)'s use of federated learning, DP-Follow-the-Regularized-Leader (DP-FTRL) algorithm, and secure aggregation techniques to train machine learning (ML) models for suggestion, prediction and correction intelligence from many users' typing data. Gboard's investment in those privacy technologies allows users' typing data to be processed locally on device, to be aggregated as early as possible, and to have strong anonymization and differential privacy where possible. Technical strategies and practices have been established to allow ML models to be trained and deployed with meaningfully formal DP guarantees and high utility. The paper also looks ahead to how technologies such as trusted execution environments may be used to further improve the privacy and security of Gboard's ML models. △ Less

Submitted 26 June, 2023; originally announced June 2023.

BOOT: Data-free Distillation of Denoising Diffusion Models with Bootstrapping

Authors: Jiatao Gu, Shuangfei Zhai, Yizhe Zhang, Lingjie Liu, Josh Susskind

Abstract: Diffusion models have demonstrated excellent potential for generating diverse images. However, their performance often suffers from slow generation due to iterative denoising. Knowledge distillation has been recently proposed as a remedy that can reduce the number of inference steps to one or a few without significant quality degradation. However, existing distillation methods either require signi… ▽ More Diffusion models have demonstrated excellent potential for generating diverse images. However, their performance often suffers from slow generation due to iterative denoising. Knowledge distillation has been recently proposed as a remedy that can reduce the number of inference steps to one or a few without significant quality degradation. However, existing distillation methods either require significant amounts of offline computation for generating synthetic training data from the teacher model or need to perform expensive online learning with the help of real data. In this work, we present a novel technique called BOOT, that overcomes these limitations with an efficient data-free distillation algorithm. The core idea is to learn a time-conditioned model that predicts the output of a pre-trained diffusion model teacher given any time step. Such a model can be efficiently trained based on bootstrapping from two consecutive sampled steps. Furthermore, our method can be easily adapted to large-scale text-to-image diffusion models, which are challenging for conventional methods given the fact that the training sets are often large and difficult to access. We demonstrate the effectiveness of our approach on several benchmark datasets in the DDIM setting, achieving comparable generation quality while being orders of magnitude faster than the diffusion teacher. The text-to-image results show that the proposed approach is able to handle highly complex distributions, shedding light on more efficient generative modeling. △ Less

Submitted 8 June, 2023; originally announced June 2023.

Comments: In progress

PLANNER: Generating Diversified Paragraph via Latent Language Diffusion Model

Authors: Yizhe Zhang, Jiatao Gu, Zhuofeng Wu, Shuangfei Zhai, Josh Susskind, Navdeep Jaitly

Abstract: Autoregressive models for text sometimes generate repetitive and low-quality output because errors accumulate during the steps of generation. This issue is often attributed to exposure bias - the difference between how a model is trained, and how it is used during inference. Denoising diffusion models provide an alternative approach in which a model can revisit and revise its output. However, they… ▽ More Autoregressive models for text sometimes generate repetitive and low-quality output because errors accumulate during the steps of generation. This issue is often attributed to exposure bias - the difference between how a model is trained, and how it is used during inference. Denoising diffusion models provide an alternative approach in which a model can revisit and revise its output. However, they can be computationally expensive and prior efforts on text have led to models that produce less fluent output compared to autoregressive models, especially for longer text and paragraphs. In this paper, we propose PLANNER, a model that combines latent semantic diffusion with autoregressive generation, to generate fluent text while exercising global control over paragraphs. The model achieves this by combining an autoregressive "decoding" module with a "planning" module that uses latent diffusion to generate semantic paragraph embeddings in a coarse-to-fine manner. The proposed method is evaluated on various conditional generation tasks, and results on semantic generation, text completion and summarization show its effectiveness in generating high-quality long-form text in an efficient manner. △ Less

Submitted 22 March, 2024; v1 submitted 4 June, 2023; originally announced June 2023.

Comments: Accepted by NeurIPS 2023, code at https://github.com/apple/ml-planner

Text-to-Image Diffusion Models can be Easily Backdoored through Multimodal Data Poisoning

Authors: Shengfang Zhai, Yinpeng Dong, Qingni Shen, Shi Pu, Yuejian Fang, Hang Su

Abstract: With the help of conditioning mechanisms, the state-of-the-art diffusion models have achieved tremendous success in guided image generation, particularly in text-to-image synthesis. To gain a better understanding of the training process and potential risks of text-to-image synthesis, we perform a systematic investigation of backdoor attack on text-to-image diffusion models and propose BadT2I, a ge… ▽ More With the help of conditioning mechanisms, the state-of-the-art diffusion models have achieved tremendous success in guided image generation, particularly in text-to-image synthesis. To gain a better understanding of the training process and potential risks of text-to-image synthesis, we perform a systematic investigation of backdoor attack on text-to-image diffusion models and propose BadT2I, a general multimodal backdoor attack framework that tampers with image synthesis in diverse semantic levels. Specifically, we perform backdoor attacks on three levels of the vision semantics: Pixel-Backdoor, Object-Backdoor and Style-Backdoor. By utilizing a regularization loss, our methods efficiently inject backdoors into a large-scale text-to-image diffusion model while preserving its utility with benign inputs. We conduct empirical experiments on Stable Diffusion, the widely-used text-to-image diffusion model, demonstrating that the large-scale diffusion model can be easily backdoored within a few fine-tuning steps. We conduct additional experiments to explore the impact of different types of textual triggers, as well as the backdoor persistence during further training, providing insights for the development of backdoor defense methods. Besides, our investigation may contribute to the copyright protection of text-to-image models in the future. △ Less

Submitted 22 October, 2023; v1 submitted 6 May, 2023; originally announced May 2023.

Comments: Carmera-ready version. To appear in ACM MM 2023. Code will be released at: https://github.com/sf-zhai/BadT2I

AutoFocusFormer: Image Segmentation off the Grid

Authors: Chen Ziwen, Kaushik Patnaik, Shuangfei Zhai, Alvin Wan, Zhile Ren, Alex Schwing, Alex Colburn, Li Fuxin

Abstract: Real world images often have highly imbalanced content density. Some areas are very uniform, e.g., large patches of blue sky, while other areas are scattered with many small objects. Yet, the commonly used successive grid downsampling strategy in convolutional deep networks treats all areas equally. Hence, small objects are represented in very few spatial locations, leading to worse results in tas… ▽ More Real world images often have highly imbalanced content density. Some areas are very uniform, e.g., large patches of blue sky, while other areas are scattered with many small objects. Yet, the commonly used successive grid downsampling strategy in convolutional deep networks treats all areas equally. Hence, small objects are represented in very few spatial locations, leading to worse results in tasks such as segmentation. Intuitively, retaining more pixels representing small objects during downsampling helps to preserve important information. To achieve this, we propose AutoFocusFormer (AFF), a local-attention transformer image recognition backbone, which performs adaptive downsampling by learning to retain the most important pixels for the task. Since adaptive downsampling generates a set of pixels irregularly distributed on the image plane, we abandon the classic grid structure. Instead, we develop a novel point-based local attention block, facilitated by a balanced clustering module and a learnable neighborhood merging module, which yields representations for our point-based versions of state-of-the-art segmentation heads. Experiments show that our AutoFocusFormer (AFF) improves significantly over baseline models of similar sizes. △ Less

Submitted 25 October, 2023; v1 submitted 24 April, 2023; originally announced April 2023.

Comments: CVPR 2023

ACM Class: I.4.6; I.4.8

Control3Diff: Learning Controllable 3D Diffusion Models from Single-view Images

Authors: Jiatao Gu, Qingzhe Gao, Shuangfei Zhai, Baoquan Chen, Lingjie Liu, Josh Susskind

Abstract: Diffusion models have recently become the de-facto approach for generative modeling in the 2D domain. However, extending diffusion models to 3D is challenging due to the difficulties in acquiring 3D ground truth data for training. On the other hand, 3D GANs that integrate implicit 3D representations into GANs have shown remarkable 3D-aware generation when trained only on single-view image datasets… ▽ More Diffusion models have recently become the de-facto approach for generative modeling in the 2D domain. However, extending diffusion models to 3D is challenging due to the difficulties in acquiring 3D ground truth data for training. On the other hand, 3D GANs that integrate implicit 3D representations into GANs have shown remarkable 3D-aware generation when trained only on single-view image datasets. However, 3D GANs do not provide straightforward ways to precisely control image synthesis. To address these challenges, We present Control3Diff, a 3D diffusion model that combines the strengths of diffusion models and 3D GANs for versatile, controllable 3D-aware image synthesis for single-view datasets. Control3Diff explicitly models the underlying latent distribution (optionally conditioned on external inputs), thus enabling direct control during the diffusion process. Moreover, our approach is general and applicable to any type of controlling input, allowing us to train it with the same diffusion objective without any auxiliary supervision. We validate the efficacy of Control3Diff on standard image generation benchmarks, including FFHQ, AFHQ, and ShapeNet, using various conditioning inputs such as images, sketches, and text prompts. Please see the project website (\url{https://jiataogu.me/control3diff}) for video comparisons. △ Less

Submitted 26 October, 2023; v1 submitted 13 April, 2023; originally announced April 2023.

Comments: Accepted by 3DV24

Stabilizing Transformer Training by Preventing Attention Entropy Collapse

Authors: Shuangfei Zhai, Tatiana Likhomanenko, Etai Littwin, Dan Busbridge, Jason Ramapuram, Yizhe Zhang, Jiatao Gu, Josh Susskind

Abstract: Training stability is of great importance to Transformers. In this work, we investigate the training dynamics of Transformers by examining the evolution of the attention layers. In particular, we track the attention entropy for each attention head during the course of training, which is a proxy for model sharpness. We identify a common pattern across different architectures and tasks, where low at… ▽ More Training stability is of great importance to Transformers. In this work, we investigate the training dynamics of Transformers by examining the evolution of the attention layers. In particular, we track the attention entropy for each attention head during the course of training, which is a proxy for model sharpness. We identify a common pattern across different architectures and tasks, where low attention entropy is accompanied by high training instability, which can take the form of oscillating loss or divergence. We denote the pathologically low attention entropy, corresponding to highly concentrated attention scores, as $\textit{entropy collapse}$. As a remedy, we propose $σ$Reparam, a simple and efficient solution where we reparametrize all linear layers with spectral normalization and an additional learned scalar. We demonstrate that $σ$Reparam successfully prevents entropy collapse in the attention layers, promoting more stable training. Additionally, we prove a tight lower bound of the attention entropy, which decreases exponentially fast with the spectral norm of the attention logits, providing additional motivation for our approach. We conduct experiments with $σ$Reparam on image classification, image self-supervised learning, machine translation, speech recognition, and language modeling tasks. We show that $σ$Reparam provides stability and robustness with respect to the choice of hyperparameters, going so far as enabling training (a) a Vision Transformer {to competitive performance} without warmup, weight decay, layer normalization or adaptive optimizers; (b) deep architectures in machine translation and (c) speech recognition to competitive performance without warmup and adaptive optimizers. Code is available at \url{https://github.com/apple/ml-sigma-reparam}. △ Less

Submitted 25 July, 2023; v1 submitted 10 March, 2023; originally announced March 2023.

Journal ref: In International Conference on Machine Learning (pp. 40770-40803). PMLR. 2023

TRACT: Denoising Diffusion Models with Transitive Closure Time-Distillation

Authors: David Berthelot, Arnaud Autef, Jierui Lin, Dian Ang Yap, Shuangfei Zhai, Siyuan Hu, Daniel Zheng, Walter Talbott, Eric Gu

Abstract: Denoising Diffusion models have demonstrated their proficiency for generative sampling. However, generating good samples often requires many iterations. Consequently, techniques such as binary time-distillation (BTD) have been proposed to reduce the number of network calls for a fixed architecture. In this paper, we introduce TRAnsitive Closure Time-distillation (TRACT), a new method that extends… ▽ More Denoising Diffusion models have demonstrated their proficiency for generative sampling. However, generating good samples often requires many iterations. Consequently, techniques such as binary time-distillation (BTD) have been proposed to reduce the number of network calls for a fixed architecture. In this paper, we introduce TRAnsitive Closure Time-distillation (TRACT), a new method that extends BTD. For single step diffusion,TRACT improves FID by up to 2.4x on the same architecture, and achieves new single-step Denoising Diffusion Implicit Models (DDIM) state-of-the-art FID (7.4 for ImageNet64, 3.8 for CIFAR10). Finally we tease apart the method through extended ablations. The PyTorch implementation will be released soon. △ Less

Submitted 7 March, 2023; originally announced March 2023.

NCL: Textual Backdoor Defense Using Noise-augmented Contrastive Learning

Authors: Shengfang Zhai, Qingni Shen, Xiaoyi Chen, Weilong Wang, Cong Li, Yuejian Fang, Zhonghai Wu

Abstract: At present, backdoor attacks attract attention as they do great harm to deep learning models. The adversary poisons the training data making the model being injected with a backdoor after being trained unconsciously by victims using the poisoned dataset. In the field of text, however, existing works do not provide sufficient defense against backdoor attacks. In this paper, we propose a Noise-augme… ▽ More At present, backdoor attacks attract attention as they do great harm to deep learning models. The adversary poisons the training data making the model being injected with a backdoor after being trained unconsciously by victims using the poisoned dataset. In the field of text, however, existing works do not provide sufficient defense against backdoor attacks. In this paper, we propose a Noise-augmented Contrastive Learning (NCL) framework to defend against textual backdoor attacks when training models with untrustworthy data. With the aim of mitigating the mapping between triggers and the target label, we add appropriate noise perturbing possible backdoor triggers, augment the training dataset, and then pull homology samples in the feature space utilizing contrastive learning objective. Experiments demonstrate the effectiveness of our method in defending three types of textual backdoor attacks, outperforming the prior works. △ Less

Submitted 3 March, 2023; originally announced March 2023.

Comments: 6 pages, 5 figures. To appear in ICASSP 2023

DNG: Taxonomy Expansion by Exploring the Intrinsic Directed Structure on Non-gaussian Space

Authors: Songlin Zhai, Weiqing Wang, Yuanfang Li, Yuan Meng

Abstract: Taxonomy expansion is the process of incorporating a large number of additional nodes (i.e., "queries") into an existing taxonomy (i.e., "seed"), with the most important step being the selection of appropriate positions for each query. Enormous efforts have been made by exploring the seed's structure. However, existing approaches are deficient in their mining of structural information in two ways:… ▽ More Taxonomy expansion is the process of incorporating a large number of additional nodes (i.e., "queries") into an existing taxonomy (i.e., "seed"), with the most important step being the selection of appropriate positions for each query. Enormous efforts have been made by exploring the seed's structure. However, existing approaches are deficient in their mining of structural information in two ways: poor modeling of the hierarchical semantics and failure to capture directionality of is-a relation. This paper seeks to address these issues by explicitly denoting each node as the combination of inherited feature (i.e., structural part) and incremental feature (i.e., supplementary part). Specifically, the inherited feature originates from "parent" nodes and is weighted by an inheritance factor. With this node representation, the hierarchy of semantics in taxonomies (i.e., the inheritance and accumulation of features from "parent" to "child") could be embodied. Additionally, based on this representation, the directionality of is-a relation could be easily translated into the irreversible inheritance of features. Inspired by the Darmois-Skitovich Theorem, we implement this irreversibility by a non-Gaussian constraint on the supplementary feature. A log-likelihood learning objective is further utilized to optimize the proposed model (dubbed DNG), whereby the required non-Gaussianity is also theoretically ensured. Extensive experimental results on two real-world datasets verify the superiority of DNG relative to several strong baselines. △ Less

Submitted 21 March, 2023; v1 submitted 22 February, 2023; originally announced February 2023.

Comments: 7figures

Group Buying Recommendation Model Based on Multi-task Learning

Authors: Shuoyao Zhai, Baichuan Liu, Deqing Yang, Yanghua Xiao

Abstract: In recent years, group buying has become one popular kind of online shopping activity, thanks to its larger sales and lower unit price. Unfortunately, research seldom focuses on recommendations specifically for group buying by now. Although some recommendation models have been proposed for group recommendation, they can not be directly used to achieve real-world group buying recommendation, due to… ▽ More In recent years, group buying has become one popular kind of online shopping activity, thanks to its larger sales and lower unit price. Unfortunately, research seldom focuses on recommendations specifically for group buying by now. Although some recommendation models have been proposed for group recommendation, they can not be directly used to achieve real-world group buying recommendation, due to the essential difference between group recommendation and group buying recommendation. In this paper, we first formalize the task of group buying recommendations into two sub-tasks. Then, based on our insights into the correlations and interactions between the two sub-tasks, we propose a novel recommendation model for group buying, MGBR, built mainly with a multi-task learning module. To improve recommendation performance further, we devise some collaborative expert networks and adjusted gates in the multi-task learning module, to promote the information interaction between the two sub-tasks. Furthermore, we propose two auxiliary losses corresponding to the two sub-tasks, to refine the representation learning in our model. Our extensive experiments not only demonstrate that the augmented representations in our model result in better performance than previous recommendation models, but also justify the impacts of the specially designed components in our model. △ Less

Submitted 25 November, 2022; originally announced November 2022.

Apple of Sodom: Hidden Backdoors in Superior Sentence Embeddings via Contrastive Learning

Authors: Xiaoyi Chen, Baisong Xin, Shengfang Zhai, Shiqing Ma, Qingni Shen, Zhonghai Wu

Abstract: This paper finds that contrastive learning can produce superior sentence embeddings for pre-trained models but is also vulnerable to backdoor attacks. We present the first backdoor attack framework, BadCSE, for state-of-the-art sentence embeddings under supervised and unsupervised learning settings. The attack manipulates the construction of positive and negative pairs so that the backdoored sampl… ▽ More This paper finds that contrastive learning can produce superior sentence embeddings for pre-trained models but is also vulnerable to backdoor attacks. We present the first backdoor attack framework, BadCSE, for state-of-the-art sentence embeddings under supervised and unsupervised learning settings. The attack manipulates the construction of positive and negative pairs so that the backdoored samples have a similar embedding with the target sample (targeted attack) or the negative embedding of its clean version (non-targeted attack). By injecting the backdoor in sentence embeddings, BadCSE is resistant against downstream fine-tuning. We evaluate BadCSE on both STS tasks and other downstream tasks. The supervised non-targeted attack obtains a performance degradation of 194.86%, and the targeted attack maps the backdoored samples to the target embedding with a 97.70% success rate while maintaining the model utility. △ Less

Submitted 20 October, 2022; originally announced October 2022.

f-DM: A Multi-stage Diffusion Model via Progressive Signal Transformation

Authors: Jiatao Gu, Shuangfei Zhai, Yizhe Zhang, Miguel Angel Bautista, Josh Susskind

Abstract: Diffusion models (DMs) have recently emerged as SoTA tools for generative modeling in various domains. Standard DMs can be viewed as an instantiation of hierarchical variational autoencoders (VAEs) where the latent variables are inferred from input-centered Gaussian distributions with fixed scales and variances. Unlike VAEs, this formulation limits DMs from changing the latent spaces and learning… ▽ More Diffusion models (DMs) have recently emerged as SoTA tools for generative modeling in various domains. Standard DMs can be viewed as an instantiation of hierarchical variational autoencoders (VAEs) where the latent variables are inferred from input-centered Gaussian distributions with fixed scales and variances. Unlike VAEs, this formulation limits DMs from changing the latent spaces and learning abstract representations. In this work, we propose f-DM, a generalized family of DMs which allows progressive signal transformation. More precisely, we extend DMs to incorporate a set of (hand-designed or learned) transformations, where the transformed input is the mean of each diffusion step. We propose a generalized formulation and derive the corresponding de-noising objective with a modified sampling algorithm. As a demonstration, we apply f-DM in image generation tasks with a range of functions, including down-sampling, blurring, and learned transformations based on the encoder of pretrained VAEs. In addition, we identify the importance of adjusting the noise levels whenever the signal is sub-sampled and propose a simple rescaling recipe. f-DM can produce high-quality samples on standard image generation benchmarks like FFHQ, AFHQ, LSUN, and ImageNet with better efficiency and semantic interpretation. △ Less

Submitted 10 October, 2022; originally announced October 2022.

Comments: 28 pages, 21 figures, work in progress

PyMIC: A deep learning toolkit for annotation-efficient medical image segmentation

Authors: Guotai Wang, Xiangde Luo, Ran Gu, Shuojue Yang, Yijie Qu, Shuwei Zhai, Qianfei Zhao, Kang Li, Shaoting Zhang

Abstract: Background and Objective: Open-source deep learning toolkits are one of the driving forces for developing medical image segmentation models. Existing toolkits mainly focus on fully supervised segmentation and require full and accurate pixel-level annotations that are time-consuming and difficult to acquire for segmentation tasks, which makes learning from imperfect labels highly desired for reduci… ▽ More Background and Objective: Open-source deep learning toolkits are one of the driving forces for developing medical image segmentation models. Existing toolkits mainly focus on fully supervised segmentation and require full and accurate pixel-level annotations that are time-consuming and difficult to acquire for segmentation tasks, which makes learning from imperfect labels highly desired for reducing the annotation cost. We aim to develop a new deep learning toolkit to support annotation-efficient learning for medical image segmentation. Methods: Our proposed toolkit named PyMIC is a modular deep learning library for medical image segmentation tasks. In addition to basic components that support development of high-performance models for fully supervised segmentation, it contains several advanced components tailored for learning from imperfect annotations, such as loading annotated and unannounced images, loss functions for unannotated, partially or inaccurately annotated images, and training procedures for co-learning between multiple networks, etc. PyMIC supports development of semi-supervised, weakly supervised and noise-robust learning methods for medical image segmentation. Results: We present several illustrative medical image segmentation tasks based on PyMIC: (1) Achieving competitive performance on fully supervised learning; (2) Semi-supervised cardiac structure segmentation with only 10% training images annotated; (3) Weakly supervised segmentation using scribble annotations; and (4) Learning from noisy labels for chest radiograph segmentation. Conclusions: The PyMIC toolkit is easy to use and facilitates efficient development of medical image segmentation models with imperfect annotations. It is modular and flexible, which enables researchers to develop high-performance models with low annotation cost. The source code is available at: https://github.com/HiLab-git/PyMIC. △ Less

Submitted 4 February, 2023; v1 submitted 19 August, 2022; originally announced August 2022.

Comments: 12 pages, 6 figures

Journal ref: Computer Methods and Programs in Biomedicine, Volume 231, April 2023, 107398

PA-Seg: Learning from Point Annotations for 3D Medical Image Segmentation using Contextual Regularization and Cross Knowledge Distillation

Authors: Shuwei Zhai, Guotai Wang, Xiangde Luo, Qiang Yue, Kang Li, Shaoting Zhang

Abstract: The success of Convolutional Neural Networks (CNNs) in 3D medical image segmentation relies on massive fully annotated 3D volumes for training that are time-consuming and labor-intensive to acquire. In this paper, we propose to annotate a segmentation target with only seven points in 3D medical images, and design a two-stage weakly supervised learning framework PA-Seg. In the first stage, we emplo… ▽ More The success of Convolutional Neural Networks (CNNs) in 3D medical image segmentation relies on massive fully annotated 3D volumes for training that are time-consuming and labor-intensive to acquire. In this paper, we propose to annotate a segmentation target with only seven points in 3D medical images, and design a two-stage weakly supervised learning framework PA-Seg. In the first stage, we employ geodesic distance transform to expand the seed points to provide more supervision signal. To further deal with unannotated image regions during training, we propose two contextual regularization strategies, i.e., multi-view Conditional Random Field (mCRF) loss and Variance Minimization (VM) loss, where the first one encourages pixels with similar features to have consistent labels, and the second one minimizes the intensity variance for the segmented foreground and background, respectively. In the second stage, we use predictions obtained by the model pre-trained in the first stage as pseudo labels. To overcome noises in the pseudo labels, we introduce a Self and Cross Monitoring (SCM) strategy, which combines self-training with Cross Knowledge Distillation (CKD) between a primary model and an auxiliary model that learn from soft labels generated by each other. Experiments on public datasets for Vestibular Schwannoma (VS) segmentation and Brain Tumor Segmentation (BraTS) demonstrated that our model trained in the first stage outperformed existing state-of-the-art weakly supervised approaches by a large margin, and after using SCM for additional training, the model's performance was close to its fully supervised counterpart on the BraTS dataset. △ Less

Submitted 13 February, 2023; v1 submitted 11 August, 2022; originally announced August 2022.

Comments: 12 pages, 10 figures, 4 tables; Accepted by IEEE TMI

GAUDI: A Neural Architect for Immersive 3D Scene Generation

Authors: Miguel Angel Bautista, Pengsheng Guo, Samira Abnar, Walter Talbott, Alexander Toshev, Zhuoyuan Chen, Laurent Dinh, Shuangfei Zhai, Hanlin Goh, Daniel Ulbricht, Afshin Dehghan, Josh Susskind

Abstract: We introduce GAUDI, a generative model capable of capturing the distribution of complex and realistic 3D scenes that can be rendered immersively from a moving camera. We tackle this challenging problem with a scalable yet powerful approach, where we first optimize a latent representation that disentangles radiance fields and camera poses. This latent representation is then used to learn a generati… ▽ More We introduce GAUDI, a generative model capable of capturing the distribution of complex and realistic 3D scenes that can be rendered immersively from a moving camera. We tackle this challenging problem with a scalable yet powerful approach, where we first optimize a latent representation that disentangles radiance fields and camera poses. This latent representation is then used to learn a generative model that enables both unconditional and conditional generation of 3D scenes. Our model generalizes previous works that focus on single objects by removing the assumption that the camera pose distribution can be shared across samples. We show that GAUDI obtains state-of-the-art performance in the unconditional generative setting across multiple datasets and allows for conditional generation of 3D scenes given conditioning variables like sparse image observations or text that describes the scene. △ Less

Submitted 27 July, 2022; originally announced July 2022.

Comments: Project webpage: https://github.com/apple/ml-gaudi

Position Prediction as an Effective Pretraining Strategy

Authors: Shuangfei Zhai, Navdeep Jaitly, Jason Ramapuram, Dan Busbridge, Tatiana Likhomanenko, Joseph Yitan Cheng, Walter Talbott, Chen Huang, Hanlin Goh, Joshua Susskind

Abstract: Transformers have gained increasing popularity in a wide range of applications, including Natural Language Processing (NLP), Computer Vision and Speech Recognition, because of their powerful representational capacity. However, harnessing this representational capacity effectively requires a large amount of data, strong regularization, or both, to mitigate overfitting. Recently, the power of the Tr… ▽ More Transformers have gained increasing popularity in a wide range of applications, including Natural Language Processing (NLP), Computer Vision and Speech Recognition, because of their powerful representational capacity. However, harnessing this representational capacity effectively requires a large amount of data, strong regularization, or both, to mitigate overfitting. Recently, the power of the Transformer has been unlocked by self-supervised pretraining strategies based on masked autoencoders which rely on reconstructing masked inputs, directly, or contrastively from unmasked content. This pretraining strategy which has been used in BERT models in NLP, Wav2Vec models in Speech and, recently, in MAE models in Vision, forces the model to learn about relationships between the content in different parts of the input using autoencoding related objectives. In this paper, we propose a novel, but surprisingly simple alternative to content reconstruction~-- that of predicting locations from content, without providing positional information for it. Doing so requires the Transformer to understand the positional relationships between different parts of the input, from their content alone. This amounts to an efficient implementation where the pretext task is a classification problem among all possible positions for each input token. We experiment on both Vision and Speech benchmarks, where our approach brings improvements over strong supervised training baselines and is comparable to modern unsupervised/self-supervised pretraining methods. Our method also enables Transformers trained without position embeddings to outperform ones trained with full position information. △ Less

Submitted 15 July, 2022; originally announced July 2022.

Comments: Accepted to ICML 2022

VIP-SLAM: An Efficient Tightly-Coupled RGB-D Visual Inertial Planar SLAM

Authors: Danpeng Chen, Shuai Wang, Weijian Xie, Shangjin Zhai, Nan Wang, Hujun Bao, Guofeng Zhang

Abstract: In this paper, we propose a tightly-coupled SLAM system fused with RGB, Depth, IMU and structured plane information. Traditional sparse points based SLAM systems always maintain a mass of map points to model the environment. Huge number of map points bring us a high computational complexity, making it difficult to be deployed on mobile devices. On the other hand, planes are common structures in ma… ▽ More In this paper, we propose a tightly-coupled SLAM system fused with RGB, Depth, IMU and structured plane information. Traditional sparse points based SLAM systems always maintain a mass of map points to model the environment. Huge number of map points bring us a high computational complexity, making it difficult to be deployed on mobile devices. On the other hand, planes are common structures in man-made environment especially in indoor environments. We usually can use a small number of planes to represent a large scene. So the main purpose of this article is to decrease the high complexity of sparse points based SLAM. We build a lightweight back-end map which consists of a few planes and map points to achieve efficient bundle adjustment (BA) with an equal or better accuracy. We use homography constraints to eliminate the parameters of numerous plane points in the optimization and reduce the complexity of BA. We separate the parameters and measurements in homography and point-to-plane constraints and compress the measurements part to further effectively improve the speed of BA. We also integrate the plane information into the whole system to realize robust planar feature extraction, data association, and global consistent planar reconstruction. Finally, we perform an ablation study and compare our method with similar methods in simulation and real environment data. Our system achieves obvious advantages in accuracy and efficiency. Even if the plane parameters are involved in the optimization, we effectively simplify the back-end map by using planar structures. The global bundle adjustment is nearly 2 times faster than the sparse points based SLAM algorithm. △ Less

Submitted 3 July, 2022; originally announced July 2022.

The Slingshot Mechanism: An Empirical Study of Adaptive Optimizers and the Grokking Phenomenon

Authors: Vimal Thilak, Etai Littwin, Shuangfei Zhai, Omid Saremi, Roni Paiss, Joshua Susskind

Abstract: The grokking phenomenon as reported by Power et al. ( arXiv:2201.02177 ) refers to a regime where a long period of overfitting is followed by a seemingly sudden transition to perfect generalization. In this paper, we attempt to reveal the underpinnings of Grokking via a series of empirical studies. Specifically, we uncover an optimization anomaly plaguing adaptive optimizers at extremely late stag… ▽ More The grokking phenomenon as reported by Power et al. ( arXiv:2201.02177 ) refers to a regime where a long period of overfitting is followed by a seemingly sudden transition to perfect generalization. In this paper, we attempt to reveal the underpinnings of Grokking via a series of empirical studies. Specifically, we uncover an optimization anomaly plaguing adaptive optimizers at extremely late stages of training, referred to as the Slingshot Mechanism. A prominent artifact of the Slingshot Mechanism can be measured by the cyclic phase transitions between stable and unstable training regimes, and can be easily monitored by the cyclic behavior of the norm of the last layers weights. We empirically observe that without explicit regularization, Grokking as reported in ( arXiv:2201.02177 ) almost exclusively happens at the onset of Slingshots, and is absent without it. While common and easily reproduced in more general settings, the Slingshot Mechanism does not follow from any known optimization theories that we are aware of, and can be easily overlooked without an in depth examination. Our work points to a surprising and useful inductive bias of adaptive gradient optimizers at late stages of training, calling for a revised theoretical analysis of their origin. △ Less

Submitted 13 June, 2022; v1 submitted 9 June, 2022; originally announced June 2022.

Comments: Removed Tex formatting commands in title Title and Abstract

Kallima: A Clean-label Framework for Textual Backdoor Attacks

Authors: Xiaoyi Chen, Yinpeng Dong, Zeyu Sun, Shengfang Zhai, Qingni Shen, Zhonghai Wu

Abstract: Although Deep Neural Network (DNN) has led to unprecedented progress in various natural language processing (NLP) tasks, research shows that deep models are extremely vulnerable to backdoor attacks. The existing backdoor attacks mainly inject a small number of poisoned samples into the training dataset with the labels changed to the target one. Such mislabeled samples would raise suspicion upon hu… ▽ More Although Deep Neural Network (DNN) has led to unprecedented progress in various natural language processing (NLP) tasks, research shows that deep models are extremely vulnerable to backdoor attacks. The existing backdoor attacks mainly inject a small number of poisoned samples into the training dataset with the labels changed to the target one. Such mislabeled samples would raise suspicion upon human inspection, potentially revealing the attack. To improve the stealthiness of textual backdoor attacks, we propose the first clean-label framework Kallima for synthesizing mimesis-style backdoor samples to develop insidious textual backdoor attacks. We modify inputs belonging to the target class with adversarial perturbations, making the model rely more on the backdoor trigger. Our framework is compatible with most existing backdoor triggers. The experimental results on three benchmark datasets demonstrate the effectiveness of the proposed method. △ Less

Submitted 3 June, 2022; originally announced June 2022.

RCMNet: A deep learning model assists CAR-T therapy for leukemia

Authors: Ruitao Zhang, Xueying Han, Ijaz Gul, Shiyao Zhai, Ying Liu, Yongbing Zhang, Yuhan Dong, Lan Ma, Dongmei Yu, Jin Zhou, Peiwu Qin

Abstract: Acute leukemia is a type of blood cancer with a high mortality rate. Current therapeutic methods include bone marrow transplantation, supportive therapy, and chemotherapy. Although a satisfactory remission of the disease can be achieved, the risk of recurrence is still high. Therefore, novel treatments are demanding. Chimeric antigen receptor-T (CAR-T) therapy has emerged as a promising approach t… ▽ More Acute leukemia is a type of blood cancer with a high mortality rate. Current therapeutic methods include bone marrow transplantation, supportive therapy, and chemotherapy. Although a satisfactory remission of the disease can be achieved, the risk of recurrence is still high. Therefore, novel treatments are demanding. Chimeric antigen receptor-T (CAR-T) therapy has emerged as a promising approach to treat and cure acute leukemia. To harness the therapeutic potential of CAR-T cell therapy for blood diseases, reliable cell morphological identification is crucial. Nevertheless, the identification of CAR-T cells is a big challenge posed by their phenotypic similarity with other blood cells. To address this substantial clinical challenge, herein we first construct a CAR-T dataset with 500 original microscopy images after staining. Following that, we create a novel integrated model called RCMNet (ResNet18 with CBAM and MHSA) that combines the convolutional neural network (CNN) and Transformer. The model shows 99.63% top-1 accuracy on the public dataset. Compared with previous reports, our model obtains satisfactory results for image classification. Although testing on the CAR-T cells dataset, a decent performance is observed, which is attributed to the limited size of the dataset. Transfer learning is adapted for RCMNet and a maximum of 83.36% accuracy has been achieved, which is higher than other SOTA models. The study evaluates the effectiveness of RCMNet on a big public dataset and translates it to a clinical dataset for diagnostic applications. △ Less

Submitted 6 May, 2022; originally announced May 2022.

Scribble-Supervised Medical Image Segmentation via Dual-Branch Network and Dynamically Mixed Pseudo Labels Supervision

Authors: Xiangde Luo, Minhao Hu, Wenjun Liao, Shuwei Zhai, Tao Song, Guotai Wang, Shaoting Zhang

Abstract: Medical image segmentation plays an irreplaceable role in computer-assisted diagnosis, treatment planning, and following-up. Collecting and annotating a large-scale dataset is crucial to training a powerful segmentation model, but producing high-quality segmentation masks is an expensive and time-consuming procedure. Recently, weakly-supervised learning that uses sparse annotations (points, scribb… ▽ More Medical image segmentation plays an irreplaceable role in computer-assisted diagnosis, treatment planning, and following-up. Collecting and annotating a large-scale dataset is crucial to training a powerful segmentation model, but producing high-quality segmentation masks is an expensive and time-consuming procedure. Recently, weakly-supervised learning that uses sparse annotations (points, scribbles, bounding boxes) for network training has achieved encouraging performance and shown the potential for annotation cost reduction. However, due to the limited supervision signal of sparse annotations, it is still challenging to employ them for networks training directly. In this work, we propose a simple yet efficient scribble-supervised image segmentation method and apply it to cardiac MRI segmentation. Specifically, we employ a dual-branch network with one encoder and two slightly different decoders for image segmentation and dynamically mix the two decoders' predictions to generate pseudo labels for auxiliary supervision. By combining the scribble supervision and auxiliary pseudo labels supervision, the dual-branch network can efficiently learn from scribble annotations end-to-end. Experiments on the public ACDC dataset show that our method performs better than current scribble-supervised segmentation methods and also outperforms several semi-supervised segmentation methods. △ Less

Submitted 3 March, 2022; originally announced March 2022.

Comments: 11 pages, 4 figures,code is available: https://github.com/HiLab-git/WSL4MIS.This is a comprehensive study about scribble-supervised medical image segmentation based on the ACDC dataset

Should I send this notification? Optimizing push notifications decision making by modeling the future

Authors: Conor O'Brien, Huasen Wu, Shaodan Zhai, Dalin Guo, Wenzhe Shi, Jonathan J Hunt

Abstract: Most recommender systems are myopic, that is they optimize based on the immediate response of the user. This may be misaligned with the true objective, such as creating long term user satisfaction. In this work we focus on mobile push notifications, where the long term effects of recommender system decisions can be particularly strong. For example, sending too many or irrelevant notifications may… ▽ More Most recommender systems are myopic, that is they optimize based on the immediate response of the user. This may be misaligned with the true objective, such as creating long term user satisfaction. In this work we focus on mobile push notifications, where the long term effects of recommender system decisions can be particularly strong. For example, sending too many or irrelevant notifications may annoy a user and cause them to disable notifications. However, a myopic system will always choose to send a notification since negative effects occur in the future. This is typically mitigated using heuristics. However, heuristics can be hard to reason about or improve, require retuning each time the system is changed, and may be suboptimal. To counter these drawbacks, there is significant interest in recommender systems that optimize directly for long-term value (LTV). Here, we describe a method for maximising LTV by using model-based reinforcement learning (RL) to make decisions about whether to send push notifications. We model the effects of sending a notification on the user's future behavior. Much of the prior work applying RL to maximise LTV in recommender systems has focused on session-based optimization, while the time horizon for notification decision making in this work extends over several days. We test this approach in an A/B test on a major social network. We show that by optimizing decisions about push notifications we are able to send less notifications and obtain a higher open rate than the baseline system, while generating the same level of user engagement on the platform as the existing, heuristic-based, system. △ Less

Submitted 17 February, 2022; originally announced February 2022.

Learning Representation from Neural Fisher Kernel with Low-rank Approximation

Authors: Ruixiang Zhang, Shuangfei Zhai, Etai Littwin, Josh Susskind

Abstract: In this paper, we study the representation of neural networks from the view of kernels. We first define the Neural Fisher Kernel (NFK), which is the Fisher Kernel applied to neural networks. We show that NFK can be computed for both supervised and unsupervised learning models, which can serve as a unified tool for representation extraction. Furthermore, we show that practical NFKs exhibit low-rank… ▽ More In this paper, we study the representation of neural networks from the view of kernels. We first define the Neural Fisher Kernel (NFK), which is the Fisher Kernel applied to neural networks. We show that NFK can be computed for both supervised and unsupervised learning models, which can serve as a unified tool for representation extraction. Furthermore, we show that practical NFKs exhibit low-rank structures. We then propose an efficient algorithm that computes a low rank approximation of NFK, which scales to large datasets and networks. We show that the low-rank approximation of NFKs derived from unsupervised generative models and supervised learning models gives rise to high-quality compact representations of data, achieving competitive results on a variety of machine learning tasks. △ Less

Submitted 3 February, 2022; originally announced February 2022.

Learning to Rank For Push Notifications Using Pairwise Expected Regret

Authors: Yuguang Yue, Yuanpu Xie, Huasen Wu, Haofeng Jia, Shaodan Zhai, Wenzhe Shi, Jonathan J Hunt

Abstract: Listwise ranking losses have been widely studied in recommender systems. However, new paradigms of content consumption present new challenges for ranking methods. In this work we contribute an analysis of learning to rank for personalized mobile push notifications and discuss the unique challenges this presents compared to traditional ranking problems. To address these challenges, we introduce a n… ▽ More Listwise ranking losses have been widely studied in recommender systems. However, new paradigms of content consumption present new challenges for ranking methods. In this work we contribute an analysis of learning to rank for personalized mobile push notifications and discuss the unique challenges this presents compared to traditional ranking problems. To address these challenges, we introduce a novel ranking loss based on weighting the pairwise loss between candidates by the expected regret incurred for misordering the pair. We demonstrate that the proposed method can outperform prior methods both in a simulated environment and in a production experiment on a major social network. △ Less

Submitted 19 January, 2022; originally announced January 2022.

MyoPS: A Benchmark of Myocardial Pathology Segmentation Combining Three-Sequence Cardiac Magnetic Resonance Images

Authors: Lei Li, Fuping Wu, Sihan Wang, Xinzhe Luo, Carlos Martin-Isla, Shuwei Zhai, Jianpeng Zhang, Yanfei Liu7, Zhen Zhang, Markus J. Ankenbrand, Haochuan Jiang, Xiaoran Zhang, Linhong Wang, Tewodros Weldebirhan Arega, Elif Altunok, Zhou Zhao, Feiyan Li, Jun Ma, Xiaoping Yang, Elodie Puybareau, Ilkay Oksuz, Stephanie Bricq, Weisheng Li, Kumaradevan Punithakumar, Sotirios A. Tsaftaris , et al. (7 additional authors not shown)

Abstract: Assessment of myocardial viability is essential in diagnosis and treatment management of patients suffering from myocardial infarction, and classification of pathology on myocardium is the key to this assessment. This work defines a new task of medical image analysis, i.e., to perform myocardial pathology segmentation (MyoPS) combining three-sequence cardiac magnetic resonance (CMR) images, which… ▽ More Assessment of myocardial viability is essential in diagnosis and treatment management of patients suffering from myocardial infarction, and classification of pathology on myocardium is the key to this assessment. This work defines a new task of medical image analysis, i.e., to perform myocardial pathology segmentation (MyoPS) combining three-sequence cardiac magnetic resonance (CMR) images, which was first proposed in the MyoPS challenge, in conjunction with MICCAI 2020. The challenge provided 45 paired and pre-aligned CMR images, allowing algorithms to combine the complementary information from the three CMR sequences for pathology segmentation. In this article, we provide details of the challenge, survey the works from fifteen participants and interpret their methods according to five aspects, i.e., preprocessing, data augmentation, learning strategy, model architecture and post-processing. In addition, we analyze the results with respect to different factors, in order to examine the key obstacles and explore potential of solutions, as well as to provide a benchmark for future research. We conclude that while promising results have been reported, the research is still in the early stage, and more in-depth exploration is needed before a successful application to the clinics. Note that MyoPS data and evaluation tool continue to be publicly available upon registration via its homepage (www.sdspeople.fudan.edu.cn/zhuangxiahai/0/myops20/). △ Less

Submitted 10 January, 2022; originally announced January 2022.

Robust Robotic Control from Pixels using Contrastive Recurrent State-Space Models

Authors: Nitish Srivastava, Walter Talbott, Martin Bertran Lopez, Shuangfei Zhai, Josh Susskind

Abstract: Modeling the world can benefit robot learning by providing a rich training signal for shaping an agent's latent state space. However, learning world models in unconstrained environments over high-dimensional observation spaces such as images is challenging. One source of difficulty is the presence of irrelevant but hard-to-model background distractions, and unimportant visual details of task-relev… ▽ More Modeling the world can benefit robot learning by providing a rich training signal for shaping an agent's latent state space. However, learning world models in unconstrained environments over high-dimensional observation spaces such as images is challenging. One source of difficulty is the presence of irrelevant but hard-to-model background distractions, and unimportant visual details of task-relevant entities. We address this issue by learning a recurrent latent dynamics model which contrastively predicts the next observation. This simple model leads to surprisingly robust robotic control even with simultaneous camera, background, and color distractions. We outperform alternatives such as bisimulation methods which impose state-similarity measures derived from divergence in future reward or future optimal actions. We obtain state-of-the-art results on the Distracting Control Suite, a challenging benchmark for pixel-based robotic control. △ Less

Submitted 2 December, 2021; originally announced December 2021.

Comments: NeurIPS Deep Reinforcement Learning Workshop 2021. Code can be found at https://github.com/apple/ml-core

Cross-Modal Object Tracking: Modality-Aware Representations and A Unified Benchmark

Authors: Chenglong Li, Tianhao Zhu, Lei Liu, Xiaonan Si, Zilin Fan, Sulan Zhai

Abstract: In many visual systems, visual tracking often bases on RGB image sequences, in which some targets are invalid in low-light conditions, and tracking performance is thus affected significantly. Introducing other modalities such as depth and infrared data is an effective way to handle imaging limitations of individual sources, but multi-modal imaging platforms usually require elaborate designs and ca… ▽ More In many visual systems, visual tracking often bases on RGB image sequences, in which some targets are invalid in low-light conditions, and tracking performance is thus affected significantly. Introducing other modalities such as depth and infrared data is an effective way to handle imaging limitations of individual sources, but multi-modal imaging platforms usually require elaborate designs and cannot be applied in many real-world applications at present. Near-infrared (NIR) imaging becomes an essential part of many surveillance cameras, whose imaging is switchable between RGB and NIR based on the light intensity. These two modalities are heterogeneous with very different visual properties and thus bring big challenges for visual tracking. However, existing works have not studied this challenging problem. In this work, we address the cross-modal object tracking problem and contribute a new video dataset, including 654 cross-modal image sequences with over 481K frames in total, and the average video length is more than 735 frames. To promote the research and development of cross-modal object tracking, we propose a new algorithm, which learns the modality-aware target representation to mitigate the appearance gap between RGB and NIR modalities in the tracking process. It is plug-and-play and could thus be flexibly embedded into different tracking frameworks. Extensive experiments on the dataset are conducted, and we demonstrate the effectiveness of the proposed algorithm in two representative tracking frameworks against 17 state-of-the-art tracking methods. We will release the dataset for free academic usage, dataset download link and code will be released soon. △ Less

Submitted 11 November, 2021; v1 submitted 7 November, 2021; originally announced November 2021.

Comments: In Submission