Search

Recently, introducing Tensor Decomposition (TD) methods into unsupervised feature selection (UFS) has been a rising research point. A tensor structure is beneficial for mining the relations between different modes and helps relieve the computation burden. However, while existing methods exploit TD to minimize the reconstruction error of a data tensor, they don't fully utilize the interpretable and discriminative information in the factor matrices. Moreover, most methods require domain knowledge to perform feature selection. To solve the above problems, we develop two Sparse Tensor Principal Component Analysis (STPCA) models that utilize the projection directions in the factor matrices to perform UFS. The first model extends Tucker Decomposition to a multiview sparse regression form and is transformed into several alternatively solved convex subproblems. The second model formulates a sparse version of the family of Tensor Singular Value Decomposition (T-SVDs) and is transformed into individual convex subproblems. For both models, we prove the optimal solution of each subproblem falls onto the Hermitian Positive Semidefinite Cone (HPSD). Accordingly, we design two fast algorithms based on HPSD projection and prove their convergence. According to the experimental results on two original synthetic datasets (Orbit and Array Signal) and five real-world datasets, the two proposed methods are suitable for handling different data tensor scenarios and outperform the state-of-the-art UFS methods.

User interests can be viewed over different time scales, mainly including stable long-term preferences and changing short-term intentions, and their combination facilitates the comprehensive sequential recommendation. However, existing work that focuses on different time scales of user modeling has ignored the negative effects of different time-scale noise, which hinders capturing actual user interests and cannot be resolved by conventional sequential denoising methods. In this paper, we propose a Long- and Short-term Interest Denoising Network (LSIDN), which employs different encoders and tailored denoising strategies to extract long- and short-term interests, respectively, achieving both comprehensive and robust user modeling. Specifically, we employ a session-level interest extraction and evolution strategy to avoid introducing inter-session behavioral noise into long-term interest modeling; we also adopt contrastive learning equipped with a homogeneous exchanging augmentation to alleviate the impact of unintentional behavioral noise on short-term interest modeling. Results of experiments on two public datasets show that LSIDN consistently outperforms state-of-the-art models and achieves significant robustness., Comment: 9 pages, accepted by SDM 2024

Parameter-efficient transfer learning (PETL) is a promising task, aiming to adapt the large-scale pre-trained model to downstream tasks with a relatively modest cost. However, current PETL methods struggle in compressing computational complexity and bear a heavy inference burden due to the complete forward process. This paper presents an efficient visual recognition paradigm, called Dynamic Adapter (Dyn-Adapter), that boosts PETL efficiency by subtly disentangling features in multiple levels. Our approach is simple: first, we devise a dynamic architecture with balanced early heads for multi-level feature extraction, along with adaptive training strategy. Second, we introduce a bidirectional sparsity strategy driven by the pursuit of powerful generalization ability. These qualities enable us to fine-tune efficiently and effectively: we reduce FLOPs during inference by 50%, while maintaining or even yielding higher recognition accuracy. Extensive experiments on diverse datasets and pretrained backbones demonstrate the potential of Dyn-Adapter serving as a general efficiency booster for PETL in vision recognition tasks., Comment: ECCV 2024

This report introduces the Qwen2 series, the latest addition to our large language models and large multimodal models. We release a comprehensive suite of foundational and instruction-tuned language models, encompassing a parameter range from 0.5 to 72 billion, featuring dense models and a Mixture-of-Experts model. Qwen2 surpasses most prior open-weight models, including its predecessor Qwen1.5, and exhibits competitive performance relative to proprietary models across diverse benchmarks on language understanding, generation, multilingual proficiency, coding, mathematics, and reasoning. The flagship model, Qwen2-72B, showcases remarkable performance: 84.2 on MMLU, 37.9 on GPQA, 64.6 on HumanEval, 89.5 on GSM8K, and 82.4 on BBH as a base language model. The instruction-tuned variant, Qwen2-72B-Instruct, attains 9.1 on MT-Bench, 48.1 on Arena-Hard, and 35.7 on LiveCodeBench. Moreover, Qwen2 demonstrates robust multilingual capabilities, proficient in approximately 30 languages, spanning English, Chinese, Spanish, French, German, Arabic, Russian, Korean, Japanese, Thai, Vietnamese, and more, underscoring its versatility and global reach. To foster community innovation and accessibility, we have made the Qwen2 model weights openly available on Hugging Face and ModelScope, and the supplementary materials including example code on GitHub. These platforms also include resources for quantization, fine-tuning, and deployment, facilitating a wide range of applications and research endeavors., Comment: 25 pages, 1 figure

Open-vocabulary object detection focusing on detecting novel categories guided by natural language. In this report, we propose Open-Vocabulary Light-Weighted Detection Transformer (OVLW-DETR), a deployment friendly open-vocabulary detector with strong performance and low latency. Building upon OVLW-DETR, we provide an end-to-end training recipe that transferring knowledge from vision-language model (VLM) to object detector with simple alignment. We align detector with the text encoder from VLM by replacing the fixed classification layer weights in detector with the class-name embeddings extracted from the text encoder. Without additional fusing module, OVLW-DETR is flexible and deployment friendly, making it easier to implement and modulate. improving the efficiency of interleaved attention computation. Experimental results demonstrate that the proposed approach is superior over existing real-time open-vocabulary detectors on standard Zero-Shot LVIS benchmark. Source code and pre-trained models are available at [https://github.com/Atten4Vis/LW-DETR]., Comment: 4 pages

Matrix-valued time series data are widely available in various applications, attracting increasing attention in the literature. However, while nonlinearity has been recognized, the literature has so far neglected a deeper and more intricate level of nonlinearity, namely the {\it row-level} nonlinear dynamics and the {\it column-level} nonlinear dynamics, which are often observed in economic and financial data. In this paper, we propose a novel two-way threshold matrix autoregression (TWTMAR) model. This model is designed to effectively characterize the threshold structure in both rows and columns of matrix-valued time series. Unlike existing models that consider a single threshold variable or assume a uniform structure change across the matrix, the TWTMAR model allows for distinct threshold effects for rows and columns using two threshold variables. This approach achieves greater dimension reduction and yields better interpretation compared to existing methods. Moreover, we propose a parameter estimation procedure leveraging the intrinsic matrix structure and investigate the asymptotic properties. The efficacy and flexibility of the model are demonstrated through both simulation studies and an empirical analysis of the Fama-French Portfolio dataset.

Prompt recovery, a crucial task in natural language processing, entails the reconstruction of prompts or instructions that language models use to convert input text into a specific output. Although pivotal, the design and effectiveness of prompts represent a challenging and relatively untapped field within NLP research. This paper delves into an exhaustive investigation of prompt recovery methodologies, employing a spectrum of pre-trained language models and strategies. Our study is a comparative analysis aimed at gauging the efficacy of various models on a benchmark dataset, with the goal of pinpointing the most proficient approach for prompt recovery. Through meticulous experimentation and detailed analysis, we elucidate the outstanding performance of the Gemma-2b-it + Phi2 model + Pretrain. This model surpasses its counterparts, showcasing its exceptional capability in accurately reconstructing prompts for text transformation tasks. Our findings offer a significant contribution to the existing knowledge on prompt recovery, shedding light on the intricacies of prompt design and offering insightful perspectives for future innovations in text rewriting and the broader field of natural language processing.

Numerous locomotion controllers have been designed based on Reinforcement Learning (RL) to facilitate blind quadrupedal locomotion traversing challenging terrains. Nevertheless, locomotion control is still a challenging task for quadruped robots traversing diverse terrains amidst unforeseen disturbances. Recently, privileged learning has been employed to learn reliable and robust quadrupedal locomotion over various terrains based on a teacher-student architecture. However, its one-encoder structure is not adequate in addressing external force perturbations. The student policy would experience inevitable performance degradation due to the feature embedding discrepancy between the feature encoder of the teacher policy and the one of the student policy. Hence, this paper presents a privileged learning framework with multiple feature encoders and a residual policy network for robust and reliable quadruped locomotion subject to various external perturbations. The multi-encoder structure can decouple latent features from different privileged information, ultimately leading to enhanced performance of the learned policy in terms of robustness, stability, and reliability. The efficiency of the proposed feature encoding module is analyzed in depth using extensive simulation data. The introduction of the residual policy network helps mitigate the performance degradation experienced by the student policy that attempts to clone the behaviors of a teacher policy. The proposed framework is evaluated on a Unitree GO1 robot, showcasing its performance enhancement over the state-of-the-art privileged learning algorithm through extensive experiments conducted on diverse terrains. Ablation studies are conducted to illustrate the efficiency of the residual policy network., Comment: 8 pages, Accepted by IROS 2024

Comprehensive and constructive evaluation protocols play an important role in the development of sophisticated text-to-video (T2V) generation models. Existing evaluation protocols primarily focus on temporal consistency and content continuity, yet largely ignore the dynamics of video content. Dynamics are an essential dimension for measuring the visual vividness and the honesty of video content to text prompts. In this study, we propose an effective evaluation protocol, termed DEVIL, which centers on the dynamics dimension to evaluate T2V models. For this purpose, we establish a new benchmark comprising text prompts that fully reflect multiple dynamics grades, and define a set of dynamics scores corresponding to various temporal granularities to comprehensively evaluate the dynamics of each generated video. Based on the new benchmark and the dynamics scores, we assess T2V models with the design of three metrics: dynamics range, dynamics controllability, and dynamics-based quality. Experiments show that DEVIL achieves a Pearson correlation exceeding 90% with human ratings, demonstrating its potential to advance T2V generation models. Code is available at https://github.com/MingXiangL/DEVIL.

We present a framework for intuitive robot programming by non-experts, leveraging natural language prompts and contextual information from the Robot Operating System (ROS). Our system integrates large language models (LLMs), enabling non-experts to articulate task requirements to the system through a chat interface. Key features of the framework include: integration of ROS with an AI agent connected to a plethora of open-source and commercial LLMs, automatic extraction of a behavior from the LLM output and execution of ROS actions/services, support for three behavior modes (sequence, behavior tree, state machine), imitation learning for adding new robot actions to the library of possible actions, and LLM reflection via human and environment feedback. Extensive experiments validate the framework, showcasing robustness, scalability, and versatility in diverse scenarios, including long-horizon tasks, tabletop rearrangements, and remote supervisory control. To facilitate the adoption of our framework and support the reproduction of our results, we have made our code open-source. You can access it at: https://github.com/huawei-noah/HEBO/tree/master/ROSLLM., Comment: This document contains 26 pages and 13 figures

Alphas are pivotal in providing signals for quantitative trading. The industry highly values the discovery of formulaic alphas for their interpretability and ease of analysis, compared with the expressive yet overfitting-prone black-box alphas. In this work, we focus on discovering formulaic alphas. Prior studies on automatically generating a collection of formulaic alphas were mostly based on genetic programming (GP), which is known to suffer from the problems of being sensitive to the initial population, converting to local optima, and slow computation speed. Recent efforts employing deep reinforcement learning (DRL) for alpha discovery have not fully addressed key practical considerations such as alpha correlations and validity, which are crucial for their effectiveness. In this work, we propose a novel framework for alpha discovery using DRL by formulating the alpha discovery process as program construction. Our agent, $\text{Alpha}^2$, assembles an alpha program optimized for an evaluation metric. A search algorithm guided by DRL navigates through the search space based on value estimates for potential alpha outcomes. The evaluation metric encourages both the performance and the diversity of alphas for a better final trading strategy. Our formulation of searching alphas also brings the advantage of pre-calculation dimensional analysis, ensuring the logical soundness of alphas, and pruning the vast search space to a large extent. Empirical experiments on real-world stock markets demonstrates $\text{Alpha}^2$'s capability to identify a diverse set of logical and effective alphas, which significantly improves the performance of the final trading strategy. The code of our method is available at https://github.com/x35f/alpha2.

We propose a transfer learning method that utilizes data representations in a semiparametric regression model. Our aim is to perform statistical inference on the parameter of primary interest in the target model while accounting for potential nonlinear effects of confounding variables. We leverage knowledge from source domains, assuming that the sample size of the source data is substantially larger than that of the target data. This knowledge transfer is carried out by the sharing of data representations, predicated on the idea that there exists a set of latent representations transferable from the source to the target domain. We address model heterogeneity between the source and target domains by incorporating domain-specific parameters in their respective models. We establish sufficient conditions for the identifiability of the models and demonstrate that the estimator for the primary parameter in the target model is both consistent and asymptotically normal. These results lay the theoretical groundwork for making statistical inferences about the main effects. Our simulation studies highlight the benefits of our method, and we further illustrate its practical applications using real-world data., Comment: 42 pages, 11 figures, 5 tables

We conducted a systematic comparison of statistical methods used for the analysis of time-to-event outcomes under various proportional and nonproportional hazard (NPH) scenarios. Our study used data from recently published oncology trials to compare the Log-rank test, still by far the most widely used option, against some available alternatives, including the MaxCombo test, the Restricted Mean Survival Time Difference (dRMST) test, the Generalized Gamma Model (GGM) and the Generalized F Model (GFM). Power, type I error rate, and time-dependent bias with respect to the RMST difference, survival probability difference, and median survival time were used to evaluate and compare the performance of these methods. In addition to the real data, we simulated three hypothetical scenarios with crossing hazards chosen so that the early and late effects 'cancel out' and used them to evaluate the ability of the aforementioned methods to detect time-specific and overall treatment effects. We implemented novel metrics for assessing the time-dependent bias in treatment effect estimates to provide a more comprehensive evaluation in NPH scenarios. Recommendations under each NPH scenario are provided by examining the type I error rate, power, and time-dependent bias associated with each statistical approach.

Food is a rich and varied dimension of cultural heritage, crucial to both individuals and social groups. To bridge the gap in the literature on the often-overlooked regional diversity in this domain, we introduce FoodieQA, a manually curated, fine-grained image-text dataset capturing the intricate features of food cultures across various regions in China. We evaluate vision-language Models (VLMs) and large language models (LLMs) on newly collected, unseen food images and corresponding questions. FoodieQA comprises three multiple-choice question-answering tasks where models need to answer questions based on multiple images, a single image, and text-only descriptions, respectively. While LLMs excel at text-based question answering, surpassing human accuracy, the open-sourced VLMs still fall short by 41\% on multi-image and 21\% on single-image VQA tasks, although closed-weights models perform closer to human levels (within 10\%). Our findings highlight that understanding food and its cultural implications remains a challenging and under-explored direction.

Diffusion models are the state of the art in text-to-image generation, but their perceptual variability remains understudied. In this paper, we examine how prompts affect image variability in black-box diffusion-based models. We propose W1KP, a human-calibrated measure of variability in a set of images, bootstrapped from existing image-pair perceptual distances. Current datasets do not cover recent diffusion models, thus we curate three test sets for evaluation. Our best perceptual distance outperforms nine baselines by up to 18 points in accuracy, and our calibration matches graded human judgements 78% of the time. Using W1KP, we study prompt reusability and show that Imagen prompts can be reused for 10-50 random seeds before new images become too similar to already generated images, while Stable Diffusion XL and DALL-E 3 can be reused 50-200 times. Lastly, we analyze 56 linguistic features of real prompts, finding that the prompt's length, CLIP embedding norm, concreteness, and word senses influence variability most. As far as we are aware, we are the first to analyze diffusion variability from a visuolinguistic perspective. Our project page is at http://w1kp.com, Comment: 13 pages, 11 figures

Learning general-purpose models from diverse datasets has achieved great success in machine learning. In robotics, however, existing methods in multi-task learning are typically constrained to a single robot and workspace, while recent work such as RT-X requires a non-trivial action normalization procedure to manually bridge the gap between different action spaces in diverse environments. In this paper, we propose the visual kinematics chain as a precise and universal representation of quasi-static actions for robot learning over diverse environments, which requires no manual adjustment since the visual kinematic chains can be automatically obtained from the robot's model and camera parameters. We propose the Visual Kinematics Transformer (VKT), a convolution-free architecture that supports an arbitrary number of camera viewpoints, and that is trained with a single objective of forecasting kinematic structures through optimal point-set matching. We demonstrate the superior performance of VKT over BC transformers as a general agent on Calvin, RLBench, Open-X, and real robot manipulation tasks. Video demonstrations can be found at https://mlzxy.github.io/visual-kinetic-chain., Comment: Submitted to CoRL 2024

In this paper, we present a light-weight detection transformer, LW-DETR, which outperforms YOLOs for real-time object detection. The architecture is a simple stack of a ViT encoder, a projector, and a shallow DETR decoder. Our approach leverages recent advanced techniques, such as training-effective techniques, e.g., improved loss and pretraining, and interleaved window and global attentions for reducing the ViT encoder complexity. We improve the ViT encoder by aggregating multi-level feature maps, and the intermediate and final feature maps in the ViT encoder, forming richer feature maps, and introduce window-major feature map organization for improving the efficiency of interleaved attention computation. Experimental results demonstrate that the proposed approach is superior over existing real-time detectors, e.g., YOLO and its variants, on COCO and other benchmark datasets. Code and models are available at (https://github.com/Atten4Vis/LW-DETR).

Training an agent to adapt to specific tasks through co-optimization of morphology and control has widely attracted attention. However, whether there exists an optimal configuration and tactics for agents in a multiagent competition scenario is still an issue that is challenging to definitively conclude. In this context, we propose competitive evolution (CompetEvo), which co-evolves agents' designs and tactics in confrontation. We build arenas consisting of three animals and their evolved derivatives, placing agents with different morphologies in direct competition with each other. The results reveal that our method enables agents to evolve a more suitable design and strategy for fighting compared to fixed-morph agents, allowing them to obtain advantages in combat scenarios. Moreover, we demonstrate the amazing and impressive behaviors that emerge when confrontations are conducted under asymmetrical morphs.

The ultimate goal of generative models is to characterize the data distribution perfectly. For image generation, common metrics of visual quality (e.g., FID), and the truthlikeness of generated images to the human eyes seem to suggest that we are close to achieving it. However, through distribution classification tasks, we find that, in the eyes of classifiers parameterized by neural networks, the strongest diffusion models are still far from this goal. Specifically, classifiers consistently and effortlessly distinguish between real and generated images in various settings. Further, we observe an intriguing discrepancy: classifiers can identify differences between diffusion models with similar performance (e.g., U-ViT-H vs. DiT-XL), but struggle to differentiate between the smallest and largest models in the same family (e.g., EDM2-XS vs. EDM2-XXL), whereas humans exhibit the opposite tendency. As an explanation, our comprehensive empirical study suggests that, unlike humans, classifiers tend to classify images through edge and high-frequency components. We believe that our methodology can serve as a probe to understand how generative models work and inspire further thought on how existing models can be improved and how the abuse of such models can be prevented.

As the field of image generation rapidly advances, traditional diffusion models and those integrated with multimodal large language models (LLMs) still encounter limitations in interpreting complex prompts and preserving image consistency pre and post-editing. To tackle these challenges, we present an innovative image editing framework that employs the robust Chain-of-Thought (CoT) reasoning and localizing capabilities of multimodal LLMs to aid diffusion models in generating more refined images. We first meticulously design a CoT process comprising instruction decomposition, region localization, and detailed description. Subsequently, we fine-tune the LISA model, a lightweight multimodal LLM, using the CoT process of Multimodal LLMs and the mask of the edited image. By providing the diffusion models with knowledge of the generated prompt and image mask, our models generate images with a superior understanding of instructions. Through extensive experiments, our model has demonstrated superior performance in image generation, surpassing existing state-of-the-art models. Notably, our model exhibits an enhanced ability to understand complex prompts and generate corresponding images, while maintaining high fidelity and consistency in images before and after generation.

Learning a single universal policy that can perform a diverse set of manipulation tasks is a promising new direction in robotics. However, existing techniques are limited to learning policies that can only perform tasks that are encountered during training, and require a large number of demonstrations to learn new tasks. Humans, on the other hand, often can learn a new task from a single unannotated demonstration. In this work, we propose the Invariance-Matching One-shot Policy Learning (IMOP) algorithm. In contrast to the standard practice of learning the end-effector's pose directly, IMOP first learns invariant regions of the state space for a given task, and then computes the end-effector's pose through matching the invariant regions between demonstrations and test scenes. Trained on the 18 RLBench tasks, IMOP achieves a success rate that outperforms the state-of-the-art consistently, by 4.5% on average over the 18 tasks. More importantly, IMOP can learn a novel task from a single unannotated demonstration, and without any fine-tuning, and achieves an average success rate improvement of $11.5\%$ over the state-of-the-art on 22 novel tasks selected across nine categories. IMOP can also generalize to new shapes and learn to manipulate objects that are different from those in the demonstration. Further, IMOP can perform one-shot sim-to-real transfer using a single real-robot demonstration., Comment: RSS 2024

In the typical urban intersection scenario, both vehicles and infrastructures are equipped with visual and LiDAR sensors. By successfully integrating the data from vehicle-side and road monitoring devices, a more comprehensive and accurate environmental perception and information acquisition can be achieved. The Calibration of sensors, as an essential component of autonomous driving technology, has consistently drawn significant attention. Particularly in scenarios involving multiple sensors collaboratively perceiving and addressing localization challenges, the requirement for inter-sensor calibration becomes crucial. Recent years have witnessed the emergence of the concept of multi-end cooperation, where infrastructure captures and transmits surrounding environment information to vehicles, bolstering their perception capabilities while mitigating costs. However, this also poses technical complexities, underscoring the pressing need for diverse end calibration. Camera and LiDAR, the bedrock sensors in autonomous driving, exhibit expansive applicability. This paper comprehensively examines and analyzes the calibration of multi-end camera-LiDAR setups from vehicle, roadside, and vehicle-road cooperation perspectives, outlining their relevant applications and profound significance. Concluding with a summary, we present our future-oriented ideas and hypotheses.

Transcranial focused ultrasound can deliver energy to the brain in a minimally invasive manner for a variety of clinical applications. However, acoustic inhomogeneities within the skull cause significant wave interactions, leading to difficulties in predicting the energy delivered to the target. We present a comprehensive examination of intracranial acoustic fields generated by focused ultrasound transducers and assess the characteristics of cranial bone that affect acoustic transmission. Acoustic field maps were generated at 88 regions of interest across 10 historical and 2 Thiel-embalmed human skull specimens with sonication at frequencies of 220 kHz, 650 kHz, and 1000 kHz. The average peak pressure attenuation was $3.1 \pm 1.4$ dB, $9.0 \pm 1.7$ dB, and $14.6 \pm 4.1$ dB, respectively. The average power attenuation was $5.0 \pm 2.4$ dB, $14.9 \pm 3.2$ dB, and $24.1 \pm 6.3$ dB, respectively. The effect of skull thickness, skull density ratio, and skull curvature on intracranial peak pressure and power was investigated and linear fits produced. Our results demonstrate that Thiel-embalmed samples fall within the confidence intervals of fits for historical samples. The effects of angulation and spacing between the transducer and the skull were also investigated. Results indicate that wave superposition resulting from skull and transducer spacing could lead to a 30-40% uncertainty in peak recorded intracranial pressure., Comment: 31 pages, 8 figures

This paper enhances our comprehension of the Distributional Synthetic Control (DSC) proposed by Gunsilius (2023), focusing on its asymptotic properties. We first establish the DSC estimator's asymptotic optimality. The essence of this optimality lies in the treatment effect estimator given by DSC achieves the lowest possible squared prediction error among all potential treatment effect estimators that depend on an average of quantiles of control units. We also establish the convergence of the DSC weights when some requirements are met, as well as the convergence rate. A significant aspect of our research is that we find DSC synthesis forms an optimal weighted average, particularly in situations where it is impractical to perfectly fit the treated unit's quantiles through the weighted average of the control units' quantiles. To corroborate our theoretical insights, we provide empirical evidence derived from simulations.

Optimization of slow-time transmit sequence endows cognitive radar with the ability to suppress strong clutter in the range-Doppler domain. However, in practice, inaccurate target velocity information or random phase error would induce uncertainty about the actual target steering vector, which would in turn severely deteriorate the the performance of the slow-time matched filter. In order to solve this problem, we propose a new optimization method for slow-time transmit sequence design. The proposed method transforms the original non-convex optimization with an uncertain target steering vector into a two-step worst-case optimization problem. For each sub-problem, we develop a corresponding trust-region Riemannian optimization algorithm. By iteratively solving the two sub-problems, a sub-optimal solution can be reached without accurate information about the target steering vector. Furthermore, the convergence property of the proposed algorithms has been analyzed and detailed proof of the convergence is given. Unlike the traditional waveform optimization method, the proposed method is designed to work with an uncertain target steering vector and therefore, is more robust in practical radar systems. Numerical simulation results in different scenarios verify the effectiveness of the proposed method in suppressing the clutter and show its advantages in terms of the output signal-to-clutter plus noise ratio (SCNR) over traditional methods.

Contrastive Learning (CL) has attracted enormous attention due to its remarkable capability in unsupervised representation learning. However, recent works have revealed the vulnerability of CL to backdoor attacks: the feature extractor could be misled to embed backdoored data close to an attack target class, thus fooling the downstream predictor to misclassify it as the target. Existing attacks usually adopt a fixed trigger pattern and poison the training set with trigger-injected data, hoping for the feature extractor to learn the association between trigger and target class. However, we find that such fixed trigger design fails to effectively associate trigger-injected data with target class in the embedding space due to special CL mechanisms, leading to a limited attack success rate (ASR). This phenomenon motivates us to find a better backdoor trigger design tailored for CL framework. In this paper, we propose a bi-level optimization approach to achieve this goal, where the inner optimization simulates the CL dynamics of a surrogate victim, and the outer optimization enforces the backdoor trigger to stay close to the target throughout the surrogate CL procedure. Extensive experiments show that our attack can achieve a higher attack success rate (e.g., $99\%$ ASR on ImageNet-100) with a very low poisoning rate ($1\%$). Besides, our attack can effectively evade existing state-of-the-art defenses. Code is available at: https://github.com/SWY666/SSL-backdoor-BLTO., Comment: Accepted by ICLR 2024

Remote photoplethysmography (rPPG) technology has become increasingly popular due to its non-invasive monitoring of various physiological indicators, making it widely applicable in multimedia interaction, healthcare, and emotion analysis. Existing rPPG methods utilize multiple datasets for training to enhance the generalizability of models. However, they often overlook the underlying conflict issues across different datasets, such as (1) label conflict resulting from different phase delays between physiological signal labels and face videos at the instance level, and (2) attribute conflict stemming from distribution shifts caused by head movements, illumination changes, skin types, etc. To address this, we introduce the DOmain-HArmonious framework (DOHA). Specifically, we first propose a harmonious phase strategy to eliminate uncertain phase delays and preserve the temporal variation of physiological signals. Next, we design a harmonious hyperplane optimization that reduces irrelevant attribute shifts and encourages the model's optimization towards a global solution that fits more valid scenarios. Our experiments demonstrate that DOHA significantly improves the performance of existing methods under multiple protocols. Our code is available at https://github.com/SWY666/rPPG-DOHA., Comment: Accepted by ACM MM 2023

Student simulation in online education is important to address dynamic learning behaviors of students with diverse backgrounds. Existing simulation models based on deep learning usually need massive training data, lacking prior knowledge in educational contexts. Large language models (LLMs) may contain such prior knowledge since they are pre-trained from a large corpus. However, because student behaviors are dynamic and multifaceted with individual differences, directly prompting LLMs is not robust nor accurate enough to capture fine-grained interactions among diverse student personas, learning behaviors, and learning outcomes. This work tackles this problem by presenting a newly annotated fine-grained large-scale dataset and proposing EduAgent, a novel generative agent framework incorporating cognitive prior knowledge (i.e., theoretical findings revealed in cognitive science) to guide LLMs to first reason correlations among various behaviors and then make simulations. Our two experiments show that EduAgent could not only mimic and predict learning behaviors of real students but also generate realistic learning behaviors of virtual students without real data.

With the development of pre-trained language models, the dense retrieval models have become promising alternatives to the traditional retrieval models that rely on exact match and sparse bag-of-words representations. Different from most dense retrieval models using a bi-encoder to encode each query or document into a dense vector, the recently proposed late-interaction multi-vector models (i.e., ColBERT and COIL) achieve state-of-the-art retrieval effectiveness by using all token embeddings to represent documents and queries and modeling their relevance with a sum-of-max operation. However, these fine-grained representations may cause unacceptable storage overhead for practical search systems. In this study, we systematically analyze the matching mechanism of these late-interaction models and show that the sum-of-max operation heavily relies on the co-occurrence signals and some important words in the document. Based on these findings, we then propose several simple document pruning methods to reduce the storage overhead and compare the effectiveness of different pruning methods on different late-interaction models. We also leverage query pruning methods to further reduce the retrieval latency. We conduct extensive experiments on both in-domain and out-domain datasets and show that some of the used pruning methods can significantly improve the efficiency of these late-interaction models without substantially hurting their retrieval effectiveness., Comment: Accepted by ACM Transactions on Information Systems

GitHub workflows or GitHub CI is a popular continuous integration platform that enables developers to automate various software engineering tasks by specifying them as workflows, i.e., YAML files with a list of jobs. However, engineering valid workflows is tedious. They are also prone to severe security issues, which can result in supply chain vulnerabilities. Recent advancements in Large Language Models (LLMs) have demonstrated their effectiveness in various software development tasks. However, GitHub workflows differ from regular programs in both structure and semantics. We perform the first comprehensive study to understand the effectiveness of LLMs on five workflow-related tasks with different levels of prompts. We curated a set of $\sim$400K workflows and generated prompts with varying detail. We also fine-tuned LLMs on GitHub workflow tasks. Our evaluation of three state-of-the-art LLMs and their fine-tuned variants revealed various interesting findings on the current effectiveness and drawbacks of LLMs.

It is common for us to feel pressure in a competition environment, which arises from the desire to obtain success comparing with other individuals or opponents. Although we might get anxious under the pressure, it could also be a drive for us to stimulate our potentials to the best in order to keep up with others. Inspired by this, we propose a competitive learning framework which is able to help individual robot to acquire knowledge from the competition, fully stimulating its dynamics potential in the race. Specifically, the competition information among competitors is introduced as the additional auxiliary signal to learn advantaged actions. We further build a Multiagent-Race environment, and extensive experiments are conducted, demonstrating that robots trained in competitive environments outperform ones that are trained with SoTA algorithms in single robot environment.

How can we construct an automated debate judge to evaluate an extensive, vibrant, multi-turn debate? This task is challenging, as judging a debate involves grappling with lengthy texts, intricate argument relationships, and multi-dimensional assessments. At the same time, current research mainly focuses on short dialogues, rarely touching upon the evaluation of an entire debate. In this paper, by leveraging Large Language Models (LLMs), we propose Debatrix, which makes the analysis and assessment of multi-turn debates more aligned with majority preferences. Specifically, Debatrix features a vertical, iterative chronological analysis and a horizontal, multi-dimensional evaluation collaboration. To align with real-world debate scenarios, we introduced the PanelBench benchmark, comparing our system's performance to actual debate outcomes. The findings indicate a notable enhancement over directly using LLMs for debate evaluation. Source code and benchmark data are available online at https://github.com/ljcleo/debatrix .

Existing pre-trained vision-language models, e.g., CLIP, have demonstrated impressive zero-shot generalization capabilities in various downstream tasks. However, the performance of these models will degrade significantly when test inputs present different distributions. To this end, we explore the concept of test-time prompt tuning (TTPT), which enables the adaptation of the CLIP model to novel downstream tasks through only one step of optimization on an unsupervised objective that involves the test sample. Motivated by in-context learning within field of natural language processing (NLP), we propose In-Context Prompt Learning (InCPL) for test-time visual recognition task. InCPL involves associating a new test sample with very few or even just one labeled example as its in-context prompt. As a result, it can reliably estimate a label for the test sample, thereby facilitating the model adaptation process. InCPL first employs a token net to represent language descriptions as visual prompts that the vision encoder of a CLIP model can comprehend. Paired with in-context examples, we further propose a context-aware unsupervised loss to optimize test sample-aware visual prompts. This optimization allows a pre-trained, frozen CLIP model to be adapted to a test sample from any task using its learned adaptive prompt. Our method has demonstrated superior performance and achieved state-of-the-art results across various downstream datasets.

Mainstream DNN-based SAR-ATR methods still face issues such as easy overfitting of a few training data, high computational overhead, and poor interpretability of the black-box model. Integrating physical knowledge into DNNs to improve performance and achieve a higher level of physical interpretability becomes the key to solving the above problems. This paper begins by focusing on the electromagnetic (EM) backscattering mechanism. We extract the EM scattering (EMS) information from the complex SAR data and integrate the physical properties of the target into the network through a dual-stream framework to guide the network to learn physically meaningful and discriminative features. Specifically, one stream is the local EMS feature (LEMSF) extraction net. It is a heterogeneous graph neural network (GNN) guided by a multi-level multi-head attention mechanism. LEMSF uses the EMS information to obtain topological structure features and high-level physical semantic features. The other stream is a CNN-based global visual features (GVF) extraction net that captures the visual features of SAR pictures from the image domain. After obtaining the two-stream features, a feature fusion subnetwork is proposed to adaptively learn the fusion strategy. Thus, the two-stream features can maximize the performance. Furthermore, the loss function is designed based on the graph distance measure to promote intra-class aggregation. We discard overly complex design ideas and effectively control the model size while maintaining algorithm performance. Finally, to better validate the performance and generalizability of the algorithms, two more rigorous evaluation protocols, namely once-for-all (OFA) and less-for-more (LFM), are used to verify the superiority of the proposed algorithm on the MSTAR.

A learning management system streamlines the management of the teaching process in a centralized place, recording, tracking, and reporting the delivery of educational courses and student performance. Educational knowledge discovery from such an e-learning system plays a crucial role in rule regulation, policy establishment, and system development. However, existing LMSs do not have embedded mining modules to directly extract knowledge. As educational modes become more complex, educational data mining efficiency from those heterogeneous student learning behaviours is gradually degraded. Therefore, an LMS incorporated with an advanced educational mining module is proposed in this study, as a means to mine efficiently from student performance records to provide valuable insights for educators in helping plan effective learning pedagogies, improve curriculum design, and guarantee quality of teaching. Through two illustrative case studies, experimental results demonstrate increased mining efficiency of the proposed mining module without information loss compared to classic educational mining algorithms. The mined knowledge reveals a set of attributes that significantly impact student academic performance, and further classification evaluation validates the identified attributes. The design and application of such an effective LMS can enable educators to learn from past student performance experiences, empowering them to guide and intervene with students in time, and eventually improve their academic success.

We present ZSL-RPPO, an improved zero-shot learning architecture that overcomes the limitations of teacher-student neural networks and enables generating robust, reliable, and versatile locomotion for quadrupedal robots in challenging terrains. We propose a new algorithm RPPO (Recurrent Proximal Policy Optimization) that directly trains recurrent neural network in partially observable environments and results in more robust training using domain randomization. Our locomotion controller supports extensive perturbation across simulation-to-reality transfer for both intrinsic and extrinsic physical parameters without further fine-tuning. This can avoid the significant decline of student's performance during simulation-to-reality transfer and therefore enhance the robustness and generalization of the locomotion controller. We deployed our controller on the Unitree A1 and Aliengo robots in real environment and exteroceptive perception is provided by either a solid-state Lidar or a depth camera. Our locomotion controller was tested in various challenging terrains like slippery surfaces, Grassy Terrain, and stairs. Our experiment results and comparison show that our approach significantly outperforms the state-of-the-art.

In this paper, we propose a novel Visual Reference Prompt (VRP) encoder that empowers the Segment Anything Model (SAM) to utilize annotated reference images as prompts for segmentation, creating the VRP-SAM model. In essence, VRP-SAM can utilize annotated reference images to comprehend specific objects and perform segmentation of specific objects in target image. It is note that the VRP encoder can support a variety of annotation formats for reference images, including \textbf{point}, \textbf{box}, \textbf{scribble}, and \textbf{mask}. VRP-SAM achieves a breakthrough within the SAM framework by extending its versatility and applicability while preserving SAM's inherent strengths, thus enhancing user-friendliness. To enhance the generalization ability of VRP-SAM, the VRP encoder adopts a meta-learning strategy. To validate the effectiveness of VRP-SAM, we conducted extensive empirical studies on the Pascal and COCO datasets. Remarkably, VRP-SAM achieved state-of-the-art performance in visual reference segmentation with minimal learnable parameters. Furthermore, VRP-SAM demonstrates strong generalization capabilities, allowing it to perform segmentation of unseen objects and enabling cross-domain segmentation. The source code and models will be available at \url{https://github.com/syp2ysy/VRP-SAM}, Comment: Accepted by CVPR 2024; The camera-ready version

Developing policies that can adjust to non-stationary environments is essential for real-world reinforcement learning applications. However, learning such adaptable policies in offline settings, with only a limited set of pre-collected trajectories, presents significant challenges. A key difficulty arises because the limited offline data makes it hard for the context encoder to differentiate between changes in the environment dynamics and shifts in the behavior policy, often leading to context misassociations. To address this issue, we introduce a novel approach called Debiased Offline Representation for fast online Adaptation (DORA). DORA incorporates an information bottleneck principle that maximizes mutual information between the dynamics encoding and the environmental data, while minimizing mutual information between the dynamics encoding and the actions of the behavior policy. We present a practical implementation of DORA, leveraging tractable bounds of the information bottleneck principle. Our experimental evaluation across six benchmark MuJoCo tasks with variable parameters demonstrates that DORA not only achieves a more precise dynamics encoding but also significantly outperforms existing baselines in terms of performance.

We introduce MobileVLM V2, a family of significantly improved vision language models upon MobileVLM, which proves that a delicate orchestration of novel architectural design, an improved training scheme tailored for mobile VLMs, and rich high-quality dataset curation can substantially benefit VLMs' performance. Specifically, MobileVLM V2 1.7B achieves better or on-par performance on standard VLM benchmarks compared with much larger VLMs at the 3B scale. Notably, our 3B model outperforms a large variety of VLMs at the 7B+ scale. Our models will be released at https://github.com/Meituan-AutoML/MobileVLM .

As a marriage between offline RL and meta-RL, the advent of offline meta-reinforcement learning (OMRL) has shown great promise in enabling RL agents to multi-task and quickly adapt while acquiring knowledge safely. Among which, Context-based OMRL (COMRL) as a popular paradigm, aims to learn a universal policy conditioned on effective task representations. In this work, by examining several key milestones in the field of COMRL, we propose to integrate these seemingly independent methodologies into a unified information theoretic framework. Most importantly, we show that the pre-existing COMRL algorithms are essentially optimizing the same mutual information objective between the task variable $\boldsymbol{M}$ and its latent representation $\boldsymbol{Z}$ by implementing various approximate bounds. Based on the theoretical insight and the information bottleneck principle, we arrive at a novel algorithm dubbed UNICORN, which exhibits remarkable generalization across a broad spectrum of RL benchmarks, context shift scenarios, data qualities and deep learning architectures, attaining the new state-of-the-art. We believe that our framework could open up avenues for new optimality bounds and COMRL algorithms., Comment: 20 pages, 8 figures, 5 tables. TLDR: We propose a novel information theoretic framework of the context-based offline meta-RL paradigm, which unifies several mainstream methods and leads to a general and state-of-the-art algorithm called UNICORN

In high-dimensional time series, the component processes are often assembled into a matrix to display their interrelationship. We focus on detecting mean shifts with unknown change point locations in these matrix time series. Series that are activated by a change may cluster along certain rows (columns), which forms mode-specific change point alignment. Leveraging mode-specific change point alignments may substantially enhance the power for change point detection. Yet, there may be no mode-specific alignments in the change point structure. We propose a powerful test to detect mode-specific change points, yet robust to non-mode-specific changes. We show the validity of using the multiplier bootstrap to compute the p-value of the proposed methods, and derive non-asymptotic bounds on the size and power of the tests. We also propose a parallel bootstrap, a computationally efficient approach for computing the p-value of the proposed adaptive test. In particular, we show the consistency of the proposed test, under mild regularity conditions. To obtain the theoretical results, we derive new, sharp bounds on Gaussian approximation and multiplier bootstrap approximation, which are of independent interest for high dimensional problems with diverging sparsity.