检索结果-内蒙古大学图书馆

Video2Reward: Generating Reward Function from Videos for Legged Robot Behavior Learning

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Zeng, Runhao Zhou, Dingjie Liang, Qiwei Liu, Junlin Li, Hui Huang, Changxin Li, Jianqiang Hu, Xiping Sun, Fuchun Artificial Intelligence Research Institute Shenzhen MSU-BIT University China College of Mechatronics and Control Engineering Shenzhen University China College of Computer Science and Software Engineering Shenzhen University China National Engineering Laboratory for Big Data System Computing Technology Shenzhen University China Department of Computer Science and Technology Tsinghua University China

Learning behavior in legged robots presents a significant challenge due to its inherent instability and complex constraints. Recent research has proposed the use of a large language model (LLM) to generate reward functions in reinforcement learning, thereby replacing the need for manually designed rewards by experts. However, this approach, which relies on textual descriptions to define learning objectives, fails to achieve controllable and precise behavior learning with clear directionality. In this paper, we introduce a new video2reward method, which directly generates reward functions from videos depicting the behaviors to be mimicked and learned. Specifically, we first process videos containing the target behaviors, converting the motion information of individuals in the videos into keypoint trajectories represented as coordinates through a video2text transforming module. These trajectories are then fed into an LLM to generate the reward function, which in turn is used to train the policy. To enhance the quality of the reward function, we develop a video-assisted iterative reward refinement scheme that visually assesses the learned behaviors and provides textual feedback to the LLM. This feedback guides the LLM to continually refine the reward function, ultimately facilitating more efficient behavior learning. Experimental results on tasks involving bipedal and quadrupedal robot motion control demonstrate that our method surpasses the performance of state-of-the-art LLM-based reward generation methods by over 37.6% in terms of human normalized score. More importantly, by switching video inputs, we find our method can rapidly learn diverse motion behaviors such as walking and running. Copyright © 2024, The Authors. All rights reserved.

关键词： Robots

A Hybrid Neural Coding Approach for Pattern Recognition with Spiking Neural Networks

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Chen, Xinyi Yang, Qu Wu, Jibin Li, Haizhou Tan, Kay Chen Department of Computing The Hong Kong Polytechnic University Hong Kong Department of Electrical and Computer Engineering National University of Singapore Singapore Shenzhen Research Institute of Big Data School of Data Science The Chinese University of Hong Kong Shenzhen [CUHK- Shenzhen China

Recently, brain-inspired spiking neural networks (SNNs) have demonstrated promising capabilities in solving pattern recognition tasks. However, these SNNs are grounded on homogeneous neurons that utilize a uniform neural coding for information representation. Given that each neural coding scheme possesses its own merits and drawbacks, these SNNs encounter challenges in achieving optimal performance such as accuracy, response time, efficiency, and robustness, all of which are crucial for practical applications. In this study, we argue that SNN architectures should be holistically designed to incorporate heterogeneous coding schemes. As an initial exploration in this direction, we propose a hybrid neural coding and learning framework, which encompasses a neural coding zoo with diverse neural coding schemes discovered in neuroscience. Additionally, it incorporates a flexible neural coding assignment strategy to accommodate task-specific requirements, along with novel layer-wise learning methods to effectively implement hybrid coding SNNs. We demonstrate the superiority of the proposed framework on image classification and sound localization tasks. Specifically, the proposed hybrid coding SNNs achieve comparable accuracy to state-of-the-art SNNs, while exhibiting significantly reduced inference latency and energy consumption, as well as high noise robustness. This study yields valuable insights into hybrid neural coding designs, paving the way for developing high-performance neuromorphic systems. Copyright © 2023, The Authors. All rights reserved.

关键词： Pattern recognition

Rethinking the visual cues in audio-visual speaker extraction

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Li, Junjie Ge, Meng Pan, Zexu Cao, Rui Wang, Longbiao Dang, Jianwu Zhang, Shiliang Tianjin Key Laboratory of Cognitive Computing and Application College of Intelligence and Computing Tianjin University Tianjin China Department of Electrical and Computer Engineering National University of Singapore Singapore Institute of Data Science National University of Singapore Singapore Shenzhen Research Institute of Big Data Shenzhen China

The Audio-Visual Speaker Extraction (AVSE) algorithm employs parallel video recording to leverage two visual cues, namely speaker identity and synchronization, to enhance performance compared to audio-only algorithms. However, the visual front-end in AVSE is often derived from a pre-trained model or end-to-end trained, making it unclear which visual cue contributes more to the speaker extraction performance. This raises the question of how to better utilize visual cues. To address this issue, we propose two training strategies that decouple the learning of the two visual cues. Our experimental results demonstrate that both visual cues are useful, with the synchronization cue having a higher impact. We introduce a more explainable model, the Decoupled Audio-Visual Speaker Extraction (DAVSE) model, which leverages both visual cues. Copyright © 2023, The Authors. All rights reserved.

关键词： Extraction

Pre-training Everywhere: Parameter-Efficient Fine-Tuning for Medical Image Analysis via Target Parameter Pre-training

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Lei, Xingliang Ye, Yiwen Chen, Ziyang Shu, Minglei Xia, Yong National Engineering Laboratory for Integrated Aero-Space-Ground-Ocean Big Data Application Technology School of Computer Science and Engineering Northwestern Polytechnical University China China Ningbo Institute of Northwestern Polytechnical University Ningbo315048 China Research & Development Institute of Northwestern Polytechnical University in Shenzhen Shenzhen518057 China

Parameter-efficient fine-tuning (PEFT) techniques have emerged to address issues of overfitting and high computational costs associated with fully fine-tuning in the paradigm of self-supervised learning. Mainstream methods based on PEFT involve adding a few trainable parameters while keeping the pre-trained parameters of the backbone fixed. These methods achieve comparative, and often superior, performance to fully fine-tuning, demonstrating the powerful representation ability of the pre-trained backbone. Despite its success, these methods typically ignore the initialization of the new parameters, often relying solely on random initialization. We argue that if pre-training is significantly beneficial, it should be applied to all parameters requiring representational capacity. Motivated by this insight, we propose a simple yet effective fine-tuning framework based on Target Parameter Pre-training (TPP). The target parameters refer to the new parameters introduced during fine-tuning. TPP includes an additional stage before PEFT to pre-train these target parameters. During this stage, the pre-trained backbone parameters are frozen, and only the target parameters are trainable. A defined pre-text task is used to encourage the target parameters to learn specific representations of downstream data. When PEFT is subsequently employed, the pre-trained target parameters are loaded to enhance fine-tuning efficiency. The proposed TPP framework is versatile, allowing for the integration of various pretext tasks for pre-training and supporting different PEFT methods as backbones. We evaluated the fine-tining performance of our method using five public datasets, including three modalities and two task types. The results demonstrate that the proposed TPP can be easily integrated into existing PEFT methods, significantly improving performance. Copyright © 2024, The Authors. All rights reserved.

关键词： Self-supervised learning

Attention Field Size Alters Patterns of Population Receptive Fields in the Early Visual Cortex

学校读者我要写书评

暂无评论

Neuroscience Bulletin 2022年第2期38卷 205-208页

作者： Bo Liu Xiaochun Wang Le Wang Qiaojun Qu Wei Zhang Bin Wang Jie Xiang Ting Yan Tianyi Yan Hui Zhang College of Medical Imaging Shanxi Medical UniversityTaiyuan 030001China Department of Radiology First Clinical Medical CollegeShanxi Medical UniversityTaiyuan 030001China College of Information and Computer Taiyuan University of TechnologyTaiyuan 030001China Intelligent Imaging Big Data and Functional Nano-imaging Engineering Research Center of Shanxi Province Shanxi Medical UniversityTaiyuan 030001China Translational Medicine Research Center Shanxi Medical UniversityTaiyuan 030001China School of Life Science Beijing Institute of TechnologyBeijing 100081China Ministry of Industry and Information Technology Key Laboratory of Convergence Medical Engineering System and Healthcare TechnologyBeijing Institute of TechnologyBeijing 100081China

Dear Editor,Visual spatial attention is a cognitive process by which prior information about the relevance of spatial locations is used to improve perceptual performance[1].Visual attention is spread across the attention field(AF),a range within visual *** AF reshapes the distribution of activity throughout the visual pathway,including the early visual cortex(EVC)and all areas along the dorsal and ventral visual cortical pathways[2,3].Importantly,the effect of attention depends not only on the spatial position of the AF but also on its size[4].

关键词： Visual Cortex visual

Traffic Aware Power Saving Communication Assisted By Double-Faced Active RIS

学校读者我要写书评

暂无评论

Traffic Aware Power Saving Communication Assisted By Double-...

IEEE International Conference on Communications (ICC)

作者： Yuyan Zhou Yang Liu Qingqing Wu Qingjiang Shi Jun Zhao School of Information and Communication Engineering Dalian University of Technology Department of Electronic Engineering Shanghai Jiao Tong University School of Software Engineering Tongji University and Shenzhen Research Institute of Big Data Shenzhen China School of Computer Science and Engineering Nanyang Technological University

Despite its high energy and hardware efficiency, some defects of the reconfigurable intelligence surface (RIS) technology have come to be realized, including the severe fading loss and restricted-to-half-space coverage. This paper proposes a novel double-faced-active (DFA)-RIS structure to overcome these defects. Besides, we utilize this novel DFA-RIS to improve power saving of the communication system. Unlike traditional power saving literature, we aim at fulfilling queueing stability and long-term power minimization in a downlink system assisted by the DFA-RIS, with a realistic data arriving process taken into consideration. Enlightened by Lyapunov control theory, we propose an online optimization strategy that adaptively adjusts DFA-RIS configuration. Each online problem can be efficiently solved by leveraging alternative directional method of multipliers (ADMM) method. Numerical results demonstrate the effectiveness of our proposed Lyapunov-guided strategy and DFA-RIS’ superiority over the classical passive RIS.

关键词：

Improving graph matching with positional reconstruction encoder-decoder network 23

学校读者我要写书评

暂无评论

Improving graph matching with positional reconstruction enco...

Proceedings of the 37th International Conference on Neural Information Processing Systems

作者： Yixiao Zhou Ruiqi Jia Hongxiang Lin Hefeng Quan Yumeng Zhao Xiaoqing Lyu Wangxuan Institute of Computer Technology Peking University School of computer science and engineering Nanjing University of Technology School of Artificial Intelligence Beijing University of Posts and Telecommunications Wangxuan Institute of Computer Technology Beijing Institute of Big Data Research Peking University

Deriving from image matching and understanding, semantic keypoint matching aims at establishing correspondence between keypoint sets in images. As graphs are powerful tools to represent points and their complex relationships, graph matching provides an effective way to find desired semantic keypoint correspondences. Recent deep graph matching methods have shown excellent performance, but there is still a lack of exploration and utilization of spatial information of keypoints as nodes in graphs. More specifically, existing methods are insufficient to capture the relative spatial relations through current graph construction approaches from the locations of semantic keypoints. To address these issues, we introduce a positional reconstruction encoder-decoder (PR-EnDec) to model intrinsic graph spatial structure, and present an end-to-end graph matching network PREGM based on PR-EnDec. Our PR-EnDec consists of a positional encoder that learns effective node spatial embedding with the affine transformation invariance, and a spatial relation decoder that further utilizes the high-order spatial information by reconstructing the locational structure of graphs contained in the node coordinates. Extensive experimental results on four public keypoint matching datasets demonstrate the effectiveness of our proposed PREGM.

关键词：

Cross-View Fuzziness and Intra-view Uncertainty-Based Weight Reconstruction for Multi-view Feature Selection 11th

学校读者我要写书评

暂无评论

Cross-View Fuzziness and Intra-view Uncertainty-Based Weigh...

11th International Joint Conference on Rough Sets, IJCRS 2025

作者： Xiang, Biao Chen, Hongmei Mi, Yong Sang, Binbin Horng, Shi-Jinn Li, Tianrui Institute of Artificial Intelligence School of Computing and Artificial Intelligence Southwest Jiaotong University Chengdu611756 China National Engineering Laboratory of Integrated Transportation Big Data Application Technology Southwest Jiaotong University Chengdu611756 China Engineering Research Center of Sustainable Urban Intelligent Transportation Ministry of Education Southwest Jiaotong University Chengdu611756 China Manufacturing Industry Chains Collaboration and Information Support Technology Key Laboratory of Sichuan Province Southwest Jiaotong University Chengdu611756 China College of Computer and Information Science Chongqing Normal University Chongqing401331 China Department of Computer Science and Information Engineering Asia University Taichung41354 Taiwan

ISBN: (纸本)9783031927430

Weight assignment is a universal and indispensable process in multi-view feature selection. However, most existing methods overlook the fuzziness and uncertainty implied in multi-view data. This may result in improper weight assignment, degrading the quality of selected features. A cross-view fuzziness and intra-view uncertainty-based weight reconstruction method (CFIUWR) is proposed for multi-view feature selection. With the fuzzy rough set theory, fuzzy approximation operators recognize cross-view fuzziness and intra-view uncertainty, which are evaluated by classification-related measurement. Cross-view fuzziness and intra-view uncertainty are incorporated to develop a reconstruction mechanism for view and feature weights. An optimization algorithm is derived, guaranteeing rapid convergence. Comparisons with state-of-the-art methods on several real-world datasets demonstrate the superiority of the CFIUWR. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.

关键词： Rough set theory

Identifying Users Transferring between Transportation Modes: A Stable Matching Approach

学校读者我要写书评

暂无评论

SSRN

SSRN 2023年

作者： Yang, Hongtai Liu, Zheng Kondor, Daniel Han, Ke He, Zhengbing School of Transportation and Logistics National Engineering Laboratory of Integrated Transportation Big Data Application Technology National United Engineering Laboratory of Integrated and Intelligent Transportation Institute of System Science and Engineering Southwest Jiaotong University Chengdu China Complexity Science Hub Vienna Austria Beijing Key Laboratory of Traffic Engineering Beijing University of Technology Beijing China

Travel data containing personal identification information needs to be anonymized before being shared and analyzed for privacy considerations. In a multi-modal transport system with different operators covering one segment of the market and collecting their own customers’ travel data, it is typically difficult to identify the trajectory of the same traveler across these different modes (e.g. shared bikes, metro, bus). This has been the main hurdle to the understanding of complete trip chains and door-to-door travel patterns. To address this challenge, this paper develops TBLink, a method to match individual travelers between different modes. The underlying idea is that if a traveler makes a transfer, the spatial-temporal signatures of the previous and next trips will be similar and must satisfy certain conditions. When this pattern happens for two travelers in the two datasets repeatedly, we can infer that the two travelers are actually the same person. Traveler matching is regarded as a stable matching problem and the Gale-Shapley algorithm is used to solve the problem. TBLink is demonstrated using metro and bus trip data from Chengdu city between January and March 2021. Results show that the precision, recall, and recovery rate of user matching are 80.81%, 92.77%, and 10.56% respectively, and the matching is more reliable as the dataset increases. Sensitivity analysis is performed to study the effect of several model parameters on the matching performance. © 2023, The Authors. All rights reserved.

关键词： Sensitivity analysis