arXiv subjects

Yang Wang

Publications and source records attributed to Yang Wang.

At least 415 records · Page 23Linked to original sources

arXiv2024

MSTF: Multiscale Transformer for Incomplete Trajectory Prediction

Motion forecasting plays a pivotal role in autonomous driving systems, enabling vehicles to execute collision warnings and rational local-path planning based on predictions of the surrounding vehicles. However, prevalent methods often assume complete observed trajectories, neglecting the potential impact of missing values induced by object occlusion, scope limitation, and sensor failures. Such oversights inevitably compromise the accuracy of trajectory predictions. To tackle this challenge, we propose an end-to-end framework, termed Multiscale Transformer (MSTF), meticulously crafted for incomplete trajectory prediction. MSTF integrates a Multiscale Attention Head (MAH) and an Information Increment-based Pattern Adaptive (IIPA) module. Specifically, the MAH component concurrently captures multiscale motion representation of trajectory sequence from various temporal granularities, utilizing a multi-head attention mechanism. This approach facilitates the modeling of global dependencies in motion across different scales, thereby mitigating the adverse effects of missing values. Additionally, the IIPA module adaptively extracts continuity representation of motion across time steps by analyzing missing patterns in the data. The continuity representation delineates motion trend at a higher level, guiding MSTF to generate predictions consistent with motion continuity. We evaluate our proposed MSTF model using two large-scale real-world datasets. Experimental results demonstrate that MSTF surpasses state-of-the-art (SOTA) models in the task of incomplete trajectory prediction, showcasing its efficacy in addressing the challenges posed by missing values in motion forecasting for autonomous driving systems.

cs.CV↗

arXiv2024

A Lightweight Video Anomaly Detection Model with Weak Supervision and Adaptive Instance Selection

Video anomaly detection is to determine whether there are any abnormal events, behaviors or objects in a given video, which enables effective and intelligent public safety management. As video anomaly labeling is both time-consuming and expensive, most existing works employ unsupervised or weakly supervised learning methods. This paper focuses on weakly supervised video anomaly detection, in which the training videos are labeled whether or not they contain any anomalies, but there is no information about which frames the anomalies are located. However, the uncertainty of weakly labeled data and the large model size prevent existing methods from wide deployment in real scenarios, especially the resource-limit situations such as edge-computing. In this paper, we develop a lightweight video anomaly detection model. On the one hand, we propose an adaptive instance selection strategy, which is based on the model's current status to select confident instances, thereby mitigating the uncertainty of weakly labeled data and subsequently promoting the model's performance. On the other hand, we design a lightweight multi-level temporal correlation attention module and an hourglass-shaped fully connected layer to construct the model, which can reduce the model parameters to only 0.56\% of the existing methods (e.g. RTFM). Our extensive experiments on two public datasets UCF-Crime and ShanghaiTech show that our model can achieve comparable or even superior AUC score compared to the state-of-the-art methods, with a significantly reduced number of model parameters.

cs.CV↗

arXiv2024

Nonuniqueness in Defining the Polarization: Nonlocal Surface Charges and the Electrostatic, Energetic, and Transport Perspectives

Ionic crystals play a central role in functional applications. Mesoscale descriptions of these crystals are based on the continuum polarization density field to represent the effective physics of charge distribution at the scale of the atomic lattice. However, a long-standing difficulty is that the classical electrostatic definition of the macroscopic polarization -- as the dipole or first moment of the charge density in a unit cell -- is not unique. This unphysical non-uniqueness has been shown to arise from starting directly with an infinite system rather than starting with a finite body and taking appropriate limits. This limit process shows that the electrostatic description requires not only the bulk polarization density, but also the surface charge density, as the effective macroscopic descriptors; that is, a nonlocal effective description. Other approaches to resolve this difficulty include relating the change in polarization to the transport of charge; or, to define the polarization as the energy-conjugate to the electric field. This work examines the relation between the classical electrostatic definition of polarization, and the transport and energy-conjugate definitions of polarization. We show the following: (1) The transport of charge does not correspond to the change in polarization in general; instead, one requires additional simplifying assumptions on the electrostatic definition of polarization for these approaches to give rise to the same macroscopic electric fields. Thus, the electrostatic definition encompasses the transport definition as a special case. (2) The energy-conjugate definition has both bulk and surface contributions; while traditional approaches neglect the surface contribution, we find that accounting for the nonlocal surface contributions is essential to obtain the correct macroscopic electric fields.

cond-mat.mes-hall↗

arXiv2024

DriveVLM: The Convergence of Autonomous Driving and Large Vision-Language Models

A primary hurdle of autonomous driving in urban environments is understanding complex and long-tail scenarios, such as challenging road conditions and delicate human behaviors. We introduce DriveVLM, an autonomous driving system leveraging Vision-Language Models (VLMs) for enhanced scene understanding and planning capabilities. DriveVLM integrates a unique combination of reasoning modules for scene description, scene analysis, and hierarchical planning. Furthermore, recognizing the limitations of VLMs in spatial reasoning and heavy computational requirements, we propose DriveVLM-Dual, a hybrid system that synergizes the strengths of DriveVLM with the traditional autonomous driving pipeline. Experiments on both the nuScenes dataset and our SUP-AD dataset demonstrate the efficacy of DriveVLM and DriveVLM-Dual in handling complex and unpredictable driving conditions. Finally, we deploy the DriveVLM-Dual on a production vehicle, verifying it is effective in real-world autonomous driving environments.

cs.CV↗

arXiv2024

The Ninth NTIRE 2024 Efficient Super-Resolution Challenge Report

This paper provides a comprehensive review of the NTIRE 2024 challenge, focusing on efficient single-image super-resolution (ESR) solutions and their outcomes. The task of this challenge is to super-resolve an input image with a magnification factor of x4 based on pairs of low and corresponding high-resolution images. The primary objective is to develop networks that optimize various aspects such as runtime, parameters, and FLOPs, while still maintaining a peak signal-to-noise ratio (PSNR) of approximately 26.90 dB on the DIV2K_LSDIR_valid dataset and 26.99 dB on the DIV2K_LSDIR_test dataset. In addition, this challenge has 4 tracks including the main track (overall performance), sub-track 1 (runtime), sub-track 2 (FLOPs), and sub-track 3 (parameters). In the main track, all three metrics (ie runtime, FLOPs, and parameter count) were considered. The ranking of the main track is calculated based on a weighted sum-up of the scores of all other sub-tracks. In sub-track 1, the practical runtime performance of the submissions was evaluated, and the corresponding score was used to determine the ranking. In sub-track 2, the number of FLOPs was considered. The score calculated based on the corresponding FLOPs was used to determine the ranking. In sub-track 3, the number of parameters was considered. The score calculated based on the corresponding parameters was used to determine the ranking. RLFN is set as the baseline for efficiency measurement. The challenge had 262 registered participants, and 34 teams made valid submissions. They gauge the state-of-the-art in efficient single-image super-resolution. To facilitate the reproducibility of the challenge and enable other researchers to build upon these findings, the code and the pre-trained model of validated solutions are made publicly available at https://github.com/Amazingren/NTIRE2024_ESR/.

cs.CV↗

arXiv2024

Hybrid Precoding With Low-Resolution PSs for Wideband Terahertz Communication Systems in The Face of Beam Squint

Terahertz (THz) communication is considered one of the most critical technologies for 6G because of its abundant bandwidth. To compensate the high propagation of THz, analog/digital hybrid precoding for THz massive multiple input multiple output (MIMO) is proposed to focus signals and extend communication range. Notably, considering hardware cost and power consumption, infinite and high-resolution phase shifters (PSs) are difficult to implement in THz massive MIMO and low-resolution PSs are typically adopted in practice. However, low-resolution PSs cause severe performance degradation. Moreover, the beam squint in wideband THz massive MIMO increases the performance degradation because of the frequency independence of the analog PSs. Motivated by the above factors, in this paper, we firstly propose a heuristic algorithm under fully connected (FC) structure, which optimize the digital precoder and the analog precoder alternately. Then we migrate the proposed heuristic algorithm to the partially-connected (PC) architecture. To further improve the performance, we extend our design to dynamic subarrays in which each RF chain is connected to any antenna that does not duplicate. The numerical results demonstrate that our proposed wideband hybrid precoding with low-resolution PSs achieves better performance to the comparisons for both FC structure and PC structure.

eess.SP↗

arXiv2024

FC3DNet: A Fully Connected Encoder-Decoder for Efficient Demoir'eing

Moiré patterns are commonly seen when taking photos of screens. Camera devices usually have limited hardware performance but take high-resolution photos. However, users are sensitive to the photo processing time, which presents a hardly considered challenge of efficiency for demoiréing methods. To balance the network speed and quality of results, we propose a \textbf{F}ully \textbf{C}onnected en\textbf{C}oder-de\textbf{C}oder based \textbf{D}emoiréing \textbf{Net}work (FC3DNet). FC3DNet utilizes features with multiple scales in each stage of the decoder for comprehensive information, which contains long-range patterns as well as various local moiré styles that both are crucial aspects in demoiréing. Besides, to make full use of multiple features, we design a Multi-Feature Multi-Attention Fusion (MFMAF) module to weigh the importance of each feature and compress them for efficiency. These designs enable our network to achieve performance comparable to state-of-the-art (SOTA) methods in real-world datasets while utilizing only a fraction of parameters, FLOPs, and runtime.

cs.CV↗

arXiv2024

Apprenticeship-Inspired Elegance: Synergistic Knowledge Distillation Empowers Spiking Neural Networks for Efficient Single-Eye Emotion Recognition

We introduce a novel multimodality synergistic knowledge distillation scheme tailored for efficient single-eye motion recognition tasks. This method allows a lightweight, unimodal student spiking neural network (SNN) to extract rich knowledge from an event-frame multimodal teacher network. The core strength of this approach is its ability to utilize the ample, coarser temporal cues found in conventional frames for effective emotion recognition. Consequently, our method adeptly interprets both temporal and spatial information from the conventional frame domain, eliminating the need for specialized sensing devices, e.g., event-based camera. The effectiveness of our approach is thoroughly demonstrated using both existing and our compiled single-eye emotion recognition datasets, achieving unparalleled performance in accuracy and efficiency over existing state-of-the-art methods.

cs.CV↗

arXiv2024

The multi-component fitting to the star formation histories in the TNG simulation

The star formation history (SFH) is a key issue in the evolution of galaxies. In this work, we developed a model based on a Gaussian and gamma function mixture to fit SFHs with varying numbers of components. Our primary objective was to use this model to reveal the shape of SFHs and the corresponding physical driving factors. Specifically, we applied this model to fit SFHs from the TNG100-1 simulation. Our study led to the following findings: 1) Our model fits with TNG star formation histories well, especially for high-mass and red galaxies; 2) A clear relationship exists between the number and shape of fitted components and the mass and color of galaxies, with notable differences observed between central/isolated and satellite galaxies. 3) Our model allowed us to extract different episodes of star formation within star formation histories with ease and analyze the duration and timing of each star formation episode. Our findings indicated a strong relationship between the timing of each star formation episode and galaxy mass and color.

astro-ph.GA↗

arXiv2024

BEV-TSR: Text-Scene Retrieval in BEV Space for Autonomous Driving

The rapid development of the autonomous driving industry has led to a significant accumulation of autonomous driving data. Consequently, there comes a growing demand for retrieving data to provide specialized optimization. However, directly applying previous image retrieval methods faces several challenges, such as the lack of global feature representation and inadequate text retrieval ability for complex driving scenes. To address these issues, firstly, we propose the BEV-TSR framework which leverages descriptive text as an input to retrieve corresponding scenes in the Bird's Eye View (BEV) space. Then to facilitate complex scene retrieval with extensive text descriptions, we employ a large language model (LLM) to extract the semantic features of the text inputs and incorporate knowledge graph embeddings to enhance the semantic richness of the language embedding. To achieve feature alignment between the BEV feature and language embedding, we propose Shared Cross-modal Embedding with a set of shared learnable embeddings to bridge the gap between these two modalities, and employ a caption generation task to further enhance the alignment. Furthermore, there lack of well-formed retrieval datasets for effective evaluation. To this end, we establish a multi-level retrieval dataset, nuScenes-Retrieval, based on the widely adopted nuScenes dataset. Experimental results on the multi-level nuScenes-Retrieval show that BEV-TSR achieves state-of-the-art performance, e.g., 85.78% and 87.66% top-1 accuracy on scene-to-text and text-to-scene retrieval respectively. Codes and datasets will be available.

cs.CV↗

arXiv2024

ELF-UA: Efficient Label-Free User Adaptation in Gaze Estimation

We consider the problem of user-adaptive 3D gaze estimation. The performance of person-independent gaze estimation is limited due to interpersonal anatomical differences. Our goal is to provide a personalized gaze estimation model specifically adapted to a target user. Previous work on user-adaptive gaze estimation requires some labeled images of the target person data to fine-tune the model at test time. However, this can be unrealistic in real-world applications, since it is cumbersome for an end-user to provide labeled images. In addition, previous work requires the training data to have both gaze labels and person IDs. This data requirement makes it infeasible to use some of the available data. To tackle these challenges, this paper proposes a new problem called efficient label-free user adaptation in gaze estimation. Our model only needs a few unlabeled images of a target user for the model adaptation. During offline training, we have some labeled source data without person IDs and some unlabeled person-specific data. Our proposed method uses a meta-learning approach to learn how to adapt to a new user with only a few unlabeled images. Our key technical innovation is to use a generalization bound from domain adaptation to define the loss function in meta-learning, so that our method can effectively make use of both the labeled source data and the unlabeled person-specific data during training. Extensive experiments validate the effectiveness of our method on several challenging benchmarks.

cs.CV↗

arXiv2024

Both Matter: Enhancing the Emotional Intelligence of Large Language Models without Compromising the General Intelligence

Emotional Intelligence (EI), consisting of emotion perception, emotion cognition and emotion expression, plays the critical roles in improving user interaction experience for the current large language model (LLM) based conversational general AI assistants. Previous works mainly focus on raising the emotion perception ability of them via naive fine-tuning on EI-related classification or regression tasks. However, this leads to the incomplete enhancement of EI and catastrophic forgetting of the general intelligence (GI). To this end, we first introduce \textsc{EiBench}, a large-scale collection of EI-related tasks in the text-to-text formation with task instructions that covers all three aspects of EI, which lays a solid foundation for the comprehensive EI enhancement of LLMs. Then a novel \underline{\textbf{Mo}}dular \underline{\textbf{E}}motional \underline{\textbf{I}}ntelligence enhancement method (\textbf{MoEI}), consisting of Modular Parameter Expansion and intra-inter modulation, is proposed to comprehensively enhance the EI of LLMs without compromise their GI. Extensive experiments on two representative LLM-based assistants, Flan-T5 and LLaMA-2-Chat, demonstrate the effectiveness of MoEI to improving EI while maintain GI.

cs.CL↗

arXiv2024

LAFMA: A Latent Flow Matching Model for Text-to-Audio Generation

Recently, the application of diffusion models has facilitated the significant development of speech and audio generation. Nevertheless, the quality of samples generated by diffusion models still needs improvement. And the effectiveness of the method is accompanied by the extensive number of sampling steps, leading to an extended synthesis time necessary for generating high-quality audio. Previous Text-to-Audio (TTA) methods mostly used diffusion models in the latent space for audio generation. In this paper, we explore the integration of the Flow Matching (FM) model into the audio latent space for audio generation. The FM is an alternative simulation-free method that trains continuous normalization flows (CNF) based on regressing vector fields. We demonstrate that our model significantly enhances the quality of generated audio samples, achieving better performance than prior models. Moreover, it reduces the number of inference steps to ten steps almost without sacrificing performance.

eess.AS↗

arXiv2024

An interpretable imbalanced semi-supervised deep learning framework for improving differential diagnosis of skin diseases

Dermatological diseases are among the most common disorders worldwide. This paper presents the first study of the interpretability and imbalanced semi-supervised learning of the multiclass intelligent skin diagnosis framework (ISDL) using 58,457 skin images with 10,857 unlabeled samples. Pseudo-labelled samples from minority classes have a higher probability at each iteration of class-rebalancing self-training, thereby promoting the utilization of unlabeled samples to solve the class imbalance problem. Our ISDL achieved a promising performance with an accuracy of 0.979, sensitivity of 0.975, specificity of 0.973, macro-F1 score of 0.974 and area under the receiver operating characteristic curve (AUC) of 0.999 for multi-label skin disease classification. The Shapley Additive explanation (SHAP) method is combined with our ISDL to explain how the deep learning model makes predictions. This finding is consistent with the clinical diagnosis. We also proposed a sampling distribution optimisation strategy to select pseudo-labelled samples in a more effective manner using ISDLplus. Furthermore, it has the potential to relieve the pressure placed on professional doctors, as well as help with practical issues associated with a shortage of such doctors in rural areas.

cs.CV↗

arXiv2024

SuperBench: Improving Cloud AI Infrastructure Reliability with Proactive Validation

Reliability in cloud AI infrastructure is crucial for cloud service providers, prompting the widespread use of hardware redundancies. However, these redundancies can inadvertently lead to hidden degradation, so called "gray failure", for AI workloads, significantly affecting end-to-end performance and concealing performance issues, which complicates root cause analysis for failures and regressions. We introduce SuperBench, a proactive validation system for AI infrastructure that mitigates hidden degradation caused by hardware redundancies and enhances overall reliability. SuperBench features a comprehensive benchmark suite, capable of evaluating individual hardware components and representing most real AI workloads. It comprises a Validator which learns benchmark criteria to clearly pinpoint defective components. Additionally, SuperBench incorporates a Selector to balance validation time and issue-related penalties, enabling optimal timing for validation execution with a tailored subset of benchmarks. Through testbed evaluation and simulation, we demonstrate that SuperBench can increase the mean time between incidents by up to 22.61x. SuperBench has been successfully deployed in Azure production, validating hundreds of thousands of GPUs over the last two years.

cs.DC↗

arXiv2024

VQPy: An Object-Oriented Approach to Modern Video Analytics

Video analytics is widely used in contemporary systems and services. At the forefront of video analytics are video queries that users develop to find objects of particular interest. Building upon the insight that video objects (e.g., human, animals, cars, etc.), the center of video analytics, are similar in spirit to objects modeled by traditional object-oriented languages, we propose to develop an object-oriented approach to video analytics. This approach, named VQPy, consists of a frontend$\unicode{x2015}$a Python variant with constructs that make it easy for users to express video objects and their interactions$\unicode{x2015}$as well as an extensible backend that can automatically construct and optimize pipelines based on video objects. We have implemented and open-sourced VQPy, which has been productized in Cisco as part of its DeepVision framework.

cs.CV↗

arXiv2024

Advancing DRL Agents in Commercial Fighting Games: Training, Integration, and Agent-Human Alignment

Deep Reinforcement Learning (DRL) agents have demonstrated impressive success in a wide range of game genres. However, existing research primarily focuses on optimizing DRL competence rather than addressing the challenge of prolonged player interaction. In this paper, we propose a practical DRL agent system for fighting games named Shūkai, which has been successfully deployed to Naruto Mobile, a popular fighting game with over 100 million registered users. Shūkai quantifies the state to enhance generalizability, introducing Heterogeneous League Training (HELT) to achieve balanced competence, generalizability, and training efficiency. Furthermore, Shūkai implements specific rewards to align the agent's behavior with human expectations. Shūkai's ability to generalize is demonstrated by its consistent competence across all characters, even though it was trained on only 13% of them. Additionally, HELT exhibits a remarkable 22% improvement in sample efficiency. Shūkai serves as a valuable training partner for players in Naruto Mobile, enabling them to enhance their abilities and skills.

cs.AI↗

arXiv2024

Characteristic time of transient response of solid oxide cells (SOCs) to changes in voltage/current: from theory to applications

The intermittency of solar and wind power can be addressed by integrating them with Solid Oxide Cells (SOCs). This study delves into the transient characteristics of SOCs and their dependence on dynamic heat and mass transfer processes. Non-dimensional analysis was used to identify influential parameters, followed by a 3-D numerical simulation-based parametric analysis to examine the dynamic gaseous and thermal responses of SOCs with varying dimensions, material properties, and operating conditions. For the first time, we proposed characteristic times to describe the relationship between SOC transients and multiple parameters. These characteristic times represent the overall heat and mass transfer rats in SOCs. Their effectiveness was validated against literature and demonstrated potential in characterizing the transient characteristics of other electrochemical cells. Besides, two examples are provided to illustrate how the characteristic times facilitate SOC design and control at minimal computational cost.

physics.flu-dyn↗