arXiv subjects

Zihan Wang

Publications and source records attributed to Zihan Wang.

At least 271 records · Page 15Linked to original sources

arXiv2024

What You See Is What Matters: A Novel Visual and Physics-Based Metric for Evaluating Video Generation Quality

As video generation models advance rapidly, assessing the quality of generated videos has become increasingly critical. Existing metrics, such as Fréchet Video Distance (FVD), Inception Score (IS), and ClipSim, measure quality primarily in latent space rather than from a human visual perspective, often overlooking key aspects like appearance and motion consistency to physical laws. In this paper, we propose a novel metric, VAMP (Visual Appearance and Motion Plausibility), that evaluates both the visual appearance and physical plausibility of generated videos. VAMP is composed of two main components: an appearance score, which assesses color, shape, and texture consistency across frames, and a motion score, which evaluates the realism of object movements. We validate VAMP through two experiments: corrupted video evaluation and generated video evaluation. In the corrupted video evaluation, we introduce various types of corruptions into real videos and measure the correlation between corruption severity and VAMP scores. In the generated video evaluation, we use state-of-the-art models to generate videos from carefully designed prompts and compare VAMP's performance to human evaluators' rankings. Our results demonstrate that VAMP effectively captures both visual fidelity and temporal consistency, offering a more comprehensive evaluation of video quality than traditional methods.

cs.CV↗

arXiv2024

I Can Tell What I am Doing: Toward Real-World Natural Language Grounding of Robot Experiences

Understanding robot behaviors and experiences through natural language is crucial for developing intelligent and transparent robotic systems. Recent advancement in large language models (LLMs) makes it possible to translate complex, multi-modal robotic experiences into coherent, human-readable narratives. However, grounding real-world robot experiences into natural language is challenging due to many reasons, such as multi-modal nature of data, differing sample rates, and data volume. We introduce RONAR, an LLM-based system that generates natural language narrations from robot experiences, aiding in behavior announcement, failure analysis, and human interaction to recover failure. Evaluated across various scenarios, RONAR outperforms state-of-the-art methods and improves failure recovery efficiency. Our contributions include a multi-modal framework for robot experience narration, a comprehensive real-robot dataset, and empirical evidence of RONAR's effectiveness in enhancing user experience in system transparency and failure analysis.

cs.RO↗

arXiv2024

SciInstruct: a Self-Reflective Instruction Annotated Dataset for Training Scientific Language Models

Large Language Models (LLMs) have shown promise in assisting scientific discovery. However, such applications are currently limited by LLMs' deficiencies in understanding intricate scientific concepts, deriving symbolic equations, and solving advanced numerical calculations. To bridge these gaps, we introduce SciInstruct, a suite of scientific instructions for training scientific language models capable of college-level scientific reasoning. Central to our approach is a novel self-reflective instruction annotation framework to address the data scarcity challenge in the science domain. This framework leverages existing LLMs to generate step-by-step reasoning for unlabelled scientific questions, followed by a process of self-reflective critic-and-revise. Applying this framework, we curated a diverse and high-quality dataset encompassing physics, chemistry, math, and formal proofs. We analyze the curated SciInstruct from multiple interesting perspectives (e.g., domain, scale, source, question type, answer length, etc.). To verify the effectiveness of SciInstruct, we fine-tuned different language models with SciInstruct, i.e., ChatGLM3 (6B and 32B), Llama3-8B-Instruct, and Mistral-7B: MetaMath, enhancing their scientific and mathematical reasoning capabilities, without sacrificing the language understanding capabilities of the base model. We release all codes and SciInstruct at https://github.com/THUDM/SciGLM.

cs.CL↗

arXiv2024

Evaluation of human-model prediction difference on the Internet Scale of Data

Evaluating models on datasets often fails to capture their behavior when faced with unexpected and diverse types of inputs. It would be beneficial if we could evaluate the difference between human annotation and model prediction for an internet number of inputs, or more generally, for an input space that enumeration is computationally impractical. Traditional model evaluation methods rely on precision and recall (PR) as metrics, which are typically estimated by comparing human annotations with model predictions on a specific dataset. This is feasible because enumerating thousands of test inputs is manageable. However, estimating PR across a large input space is challenging because enumeration becomes computationally infeasible. We propose OmniInput, a novel approach to evaluate and compare NNs by the PR of an input space. OmniInput is distinctive from previous works as its estimated PR reflects the estimation of the differences between human annotation and model prediction in the input space which is usually too huge to be enumerated. We empirically validate our method within an enumerable input space, and our experiments demonstrate that OmniInput can effectively estimate and compare precision and recall for (large) language models within a broad input space that is not enumerable.

cs.LG↗

arXiv2024

R^3AG: First Workshop on Refined and Reliable Retrieval Augmented Generation

Retrieval-augmented generation (RAG) has gained wide attention as the key component to improve generative models with external knowledge augmentation from information retrieval. It has shown great prominence in enhancing the functionality and performance of large language model (LLM)-based applications. However, with the comprehensive application of RAG, more and more problems and limitations have been identified, thus urgently requiring further fundamental exploration to improve current RAG frameworks. This workshop aims to explore in depth how to conduct refined and reliable RAG for downstream AI tasks. To this end, we propose to organize the first R3AG workshop at SIGIR-AP 2024 to call for participants to re-examine and formulate the basic principles and practical implementation of refined and reliable RAG. The workshop serves as a platform for both academia and industry researchers to conduct discussions, share insights, and foster research to build the next generation of RAG systems. Participants will engage in discussions and presentations focusing on fundamental challenges, cutting-edge research, and potential pathways to improve RAG. At the end of the workshop, we aim to have a clearer understanding of how to improve the reliability and applicability of RAG with more robust information retrieval and language generation.

cs.IR↗

arXiv2024

Effective Finite Time Stability Control for Human-Machine Shared Vehicle Following System

With the development of intelligent connected vehicle technology, human-machine shared control has gained popularity in vehicle following due to its effectiveness in driver assistance. However, traditional vehicle following systems struggle to maintain stability when driver reaction time fluctuates, as these variations require different levels of system intervention. To address this issue, the proposed human-machine shared vehicle following assistance system (HM-VFAS) integrates driver outputs under various states with the assistance system. The system employs an intelligent driver model that accounts for reaction time delays, simulating time-varying driver outputs. A control authority allocation strategy is designed to dynamically adjust the level of intervention based on real-time driver state assessment. To handle instability from driver authority switching, the proposed solution includes a two-layer adaptive finite time sliding mode controller (A-FTSMC). The first layer is an integral sliding mode adaptive controller that ensures robustness by compensating for uncertainties in the driver output. The second layer is a fast non-singular terminal sliding mode controller designed to accelerate convergence for rapid stabilization. Using real driver videos as inputs, the performance of the HM-VFAS was evaluated. Results show that the proposed control strategy maintains a safe distance under time-varying driver states, with the actual acceleration error relative to the target acceleration maintained within 0.5m/s~2 and the maximum acceleration error reduced by 1.2m/s~2. Compared to traditional controllers, the A-FTSMC controller offers faster convergence and less vibration, reducing the stabilization time by 27.3%.

eess.SY↗

arXiv2024

Imprompter: Tricking LLM Agents into Improper Tool Use

Large Language Model (LLM) Agents are an emerging computing paradigm that blends generative machine learning with tools such as code interpreters, web browsing, email, and more generally, external resources. These agent-based systems represent an emerging shift in personal computing. We contribute to the security foundations of agent-based systems and surface a new class of automatically computed obfuscated adversarial prompt attacks that violate the confidentiality and integrity of user resources connected to an LLM agent. We show how prompt optimization techniques can find such prompts automatically given the weights of a model. We demonstrate that such attacks transfer to production-level agents. For example, we show an information exfiltration attack on Mistral's LeChat agent that analyzes a user's conversation, picks out personally identifiable information, and formats it into a valid markdown command that results in leaking that data to the attacker's server. This attack shows a nearly 80% success rate in an end-to-end evaluation. We conduct a range of experiments to characterize the efficacy of these attacks and find that they reliably work on emerging agent-based systems like Mistral's LeChat, ChatGLM, and Meta's Llama. These attacks are multimodal, and we show variants in the text-only and image domains.

cs.CR↗

arXiv2024

AGSENet: A Robust Road Ponding Detection Method for Proactive Traffic Safety

Road ponding, a prevalent traffic hazard, poses a serious threat to road safety by causing vehicles to lose control and leading to accidents ranging from minor fender benders to severe collisions. Existing technologies struggle to accurately identify road ponding due to complex road textures and variable ponding coloration influenced by reflection characteristics. To address this challenge, we propose a novel approach called Self-Attention-based Global Saliency-Enhanced Network (AGSENet) for proactive road ponding detection and traffic safety improvement. AGSENet incorporates saliency detection techniques through the Channel Saliency Information Focus (CSIF) and Spatial Saliency Information Enhancement (SSIE) modules. The CSIF module, integrated into the encoder, employs self-attention to highlight similar features by fusing spatial and channel information. The SSIE module, embedded in the decoder, refines edge features and reduces noise by leveraging correlations across different feature levels. To ensure accurate and reliable evaluation, we corrected significant mislabeling and missing annotations in the Puddle-1000 dataset. Additionally, we constructed the Foggy-Puddle and Night-Puddle datasets for road ponding detection in low-light and foggy conditions, respectively. Experimental results demonstrate that AGSENet outperforms existing methods, achieving IoU improvements of 2.03\%, 0.62\%, and 1.06\% on the Puddle-1000, Foggy-Puddle, and Night-Puddle datasets, respectively, setting a new state-of-the-art in this field. Finally, we verified the algorithm's reliability on edge computing devices. This work provides a valuable reference for proactive warning research in road traffic safety.

cs.CV↗

arXiv2024

Multi-diseases detection with memristive system on chip

This study presents the first implementation of multilayer neural networks on a memristor/CMOS integrated system on chip (SoC) to simultaneously detect multiple diseases. To overcome limitations in medical data, generative AI techniques are used to enhance the dataset, improving the classifier's robustness and diversity. The system achieves notable performance with low latency, high accuracy (91.82%), and energy efficiency, facilitated by end-to-end execution on a memristor-based SoC with ten 256x256 crossbar arrays and an integrated on-chip processor. This research showcases the transformative potential of memristive in-memory computing hardware in accelerating machine learning applications for medical diagnostics.

cs.AR↗

arXiv2024

Multi-step Problem Solving Through a Verifier: An Empirical Analysis on Model-induced Process Supervision

Process supervision, using a trained verifier to evaluate the intermediate steps generated by a reasoner, has demonstrated significant improvements in multi-step problem solving. In this paper, to avoid the expensive effort of human annotation on the verifier training data, we introduce Model-induced Process Supervision (MiPS), a novel method for automating data curation. MiPS annotates an intermediate step by sampling completions of this solution through the reasoning model, and obtaining an accuracy defined as the proportion of correct completions. Inaccuracies of the reasoner would cause MiPS underestimating the accuracy of intermediate steps, therefore, we suggest and empirically show that verification focusing on high predicted scores of the verifier shall be preferred over that of low predicted scores, contrary to prior observations on human curated data. Our approach significantly improves the performance of PaLM 2 on math and coding tasks (accuracy +0.67% on GSM8K, +4.16% on MATH, +0.92% on MBPP compared with an output supervision trained verifier). Additionally, our study demonstrates that the verifier exhibits strong generalization ability across different reasoning models.

cs.AI↗

arXiv2024

Sim-to-Real Transfer via 3D Feature Fields for Vision-and-Language Navigation

Vision-and-language navigation (VLN) enables the agent to navigate to a remote location in 3D environments following the natural language instruction. In this field, the agent is usually trained and evaluated in the navigation simulators, lacking effective approaches for sim-to-real transfer. The VLN agents with only a monocular camera exhibit extremely limited performance, while the mainstream VLN models trained with panoramic observation, perform better but are difficult to deploy on most monocular robots. For this case, we propose a sim-to-real transfer approach to endow the monocular robots with panoramic traversability perception and panoramic semantic understanding, thus smoothly transferring the high-performance panoramic VLN models to the common monocular robots. In this work, the semantic traversable map is proposed to predict agent-centric navigable waypoints, and the novel view representations of these navigable waypoints are predicted through the 3D feature fields. These methods broaden the limited field of view of the monocular robots and significantly improve navigation performance in the real world. Our VLN system outperforms previous SOTA monocular VLN methods in R2R-CE and RxR-CE benchmarks within the simulation environments and is also validated in real-world environments, providing a practical and high-performance solution for real-world VLN.

cs.RO↗

arXiv2024

Commentary Generation from Data Records of Multiplayer Strategy Esports Game

Esports, a sports competition on video games, has become one of the most important sporting events. Although esports play logs have been accumulated, only a small portion of them accompany text commentaries for the audience to retrieve and understand the plays. In this study, we therefore introduce the task of generating game commentaries from esports' data records. We first build large-scale esports data-to-text datasets that pair structured data and commentaries from a popular esports game, League of Legends. We then evaluate Transformer-based models to generate game commentaries from structured data records, while examining the impact of the pre-trained language models. Evaluation results on our dataset revealed the challenges of this novel task. We will release our dataset to boost potential research in the data-to-text generation community.

cs.CL↗

arXiv2024

Wide Neural Networks Trained with Weight Decay Provably Exhibit Neural Collapse

Deep neural networks (DNNs) at convergence consistently represent the training data in the last layer via a highly symmetric geometric structure referred to as neural collapse. This empirical evidence has spurred a line of theoretical research aimed at proving the emergence of neural collapse, mostly focusing on the unconstrained features model. Here, the features of the penultimate layer are free variables, which makes the model data-agnostic and, hence, puts into question its ability to capture DNN training. Our work addresses the issue, moving away from unconstrained features and studying DNNs that end with at least two linear layers. We first prove generic guarantees on neural collapse that assume (i) low training error and balancedness of the linear layers (for within-class variability collapse), and (ii) bounded conditioning of the features before the linear part (for orthogonality of class-means, as well as their alignment with weight matrices). We then show that such assumptions hold for gradient descent training with weight decay: (i) for networks with a wide first layer, we prove low training error and balancedness, and (ii) for solutions that are either nearly optimal or stable under large learning rates, we additionally prove the bounded conditioning. Taken together, our results are the first to show neural collapse in the end-to-end training of DNNs.

cs.LG↗

arXiv2024

Gaussian Splatting in Mirrors: Reflection-Aware Rendering via Virtual Camera Optimization

Recent advancements in 3D Gaussian Splatting (3D-GS) have revolutionized novel view synthesis, facilitating real-time, high-quality image rendering. However, in scenarios involving reflective surfaces, particularly mirrors, 3D-GS often misinterprets reflections as virtual spaces, resulting in blurred and inconsistent multi-view rendering within mirrors. Our paper presents a novel method aimed at obtaining high-quality multi-view consistent reflection rendering by modelling reflections as physically-based virtual cameras. We estimate mirror planes with depth and normal estimates from 3D-GS and define virtual cameras that are placed symmetrically about the mirror plane. These virtual cameras are then used to explain mirror reflections in the scene. To address imperfections in mirror plane estimates, we propose a straightforward yet effective virtual camera optimization method to enhance reflection quality. We collect a new mirror dataset including three real-world scenarios for more diverse evaluation. Experimental validation on both Mirror-Nerf and our real-world dataset demonstrate the efficacy of our approach. We achieve comparable or superior results while significantly reducing training time compared to previous state-of-the-art.

cs.CV↗

arXiv2024

Vision-driven UAV River Following: Benchmarking with Safe Reinforcement Learning

In this study, we conduct a comprehensive benchmark of the Safe Reinforcement Learning (Safe RL) algorithms for the task of vision-driven river following of Unmanned Aerial Vehicle (UAV) in a Unity-based photo-realistic simulation environment. We empirically validate the effectiveness of semantic-augmented image encoding method, assessing its superiority based on Relative Entropy and the quality of water pixel reconstruction. The determination of the encoding dimension, guided by reconstruction loss, contributes to a more compact state representation, facilitating the training of Safe RL policies. Across all benchmarked Safe RL algorithms, we find that First Order Constrained Optimization in Policy Space achieves the optimal balance between reward acquisition and safety compliance. Notably, our results reveal that on-policy algorithms consistently outperform both off-policy and model-based counterparts in both training and testing environments. Importantly, the benchmarking outcomes and the vision encoding methodology extend beyond UAVs, and are applicable to Autonomous Surface Vehicles (ASVs) engaged in autonomous navigation in confined waters.

cs.RO↗

arXiv2024

CogVLM2: Visual Language Models for Image and Video Understanding

Beginning with VisualGLM and CogVLM, we are continuously exploring VLMs in pursuit of enhanced vision-language fusion, efficient higher-resolution architecture, and broader modalities and applications. Here we propose the CogVLM2 family, a new generation of visual language models for image and video understanding including CogVLM2, CogVLM2-Video and GLM-4V. As an image understanding model, CogVLM2 inherits the visual expert architecture with improved training recipes in both pre-training and post-training stages, supporting input resolution up to $1344 \times 1344$ pixels. As a video understanding model, CogVLM2-Video integrates multi-frame input with timestamps and proposes automated temporal grounding data construction. Notably, CogVLM2 family has achieved state-of-the-art results on benchmarks like MMBench, MM-Vet, TextVQA, MVBench and VCGBench. All models are open-sourced in https://github.com/THUDM/CogVLM2 and https://github.com/THUDM/GLM-4, contributing to the advancement of the field.

cs.CV↗

arXiv2024

Evaluating the effectiveness of public policies on COVID-19 containment: A PSM-DID approach

The implementation of public policies is crucial in controlling the spread of COVID-19. However, the effectiveness of different policies can vary across different aspects of epidemic containment. Identifying the most effective policies is essential for providing informed recommendations for pandemic control. This paper examines the relationship between various public policy responses and their impact on COVID-19 containment. Using the propensity score matching-difference in differences (PSM-DID) model to address endogeneity, we analyze the causal significance of each policy on epidemic control. Our analysis reveals that that policies related to vaccine delivery, debt relief, and the cancellation of public events are the most effective measures. These findings provide key insights for policymakers, highlighting the importance of focusing on specific, high-impact measures in managing public health crises.

stat.AP↗

arXiv2024

Reproduction of scan B-statistic for kernel change-point detection algorithm

Change-point detection has garnered significant attention due to its broad range of applications, including epidemic disease outbreaks, social network evolution, image analysis, and wireless communications. In an online setting, where new data samples arrive sequentially, it is crucial to continuously test whether these samples originate from a different distribution. Ideally, the detection algorithm should be distribution-free to ensure robustness in real-world applications. In this paper, we reproduce a recently proposed online change-point detection algorithm based on an efficient kernel-based scan B-statistic, and compare its performance with two commonly used parametric statistics. Our numerical experiments demonstrate that the scan B-statistic consistently delivers superior performance. In more challenging scenarios, parametric methods may fail to detect changes, whereas the scan B-statistic successfully identifies them in a timely manner. Additionally, the use of subsampling techniques offers a modest improvement to the original algorithm.

stat.ML↗