检索结果-内蒙古大学图书馆

Skeleton-Based Human Action Recognition with Noisy Labels

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Xu, Yi Peng, Kunyu Wen, Di Liu, Ruiping Zheng, Junwei Chen, Yufan Zhang, Jiaming Roitberg, Alina Yang, Kailun Stiefelhagen, Rainer Institute for Anthropomatics and Robotics Karlsruhe Institute of Technology Germany Institute for Artificial Intelligence University of Stuttgart Germany School of Robotics Hunan University China National Engineering Research Center of Robot Visual Perception and Control Technology Hunan University China

Understanding human actions from body poses is critical for assistive robots sharing space with humans in order to make informed and safe decisions about the next interaction. However, precise temporal localization and annotation of activity sequences is time-consuming and the resulting labels are often noisy. If not effectively addressed, label noise negatively affects the model’s training, resulting in lower recognition quality. Despite its importance, addressing label noise for skeleton-based action recognition has been overlooked so far. In this study, we bridge this gap by implementing a framework that augments well-established skeleton-based human action recognition methods with label-denoising strategies from various research areas to serve as the initial benchmark. Observations reveal that these baselines yield only marginal performance when dealing with sparse skeleton data. Consequently, we introduce a novel methodology, NoiseEraSAR, which integrates global sample selection, co-teaching, and Cross-Modal Mixture-of-Experts (CM-MOE) strategies, aimed at mitigating the adverse impacts of label noise. Our proposed approach demonstrates better performance on the established benchmark, setting new state-of-the-art standards. The source code for this study is accessible at https://***/xuyizdby/NoiseEraSAR. Copyright © 2024, The Authors. All rights reserved.

关键词： Musculoskeletal system

Domain Adaptation in visual Reinforcement Learning via Self-Expert Imitation with Purifying Latent Feature

学校读者我要写书评

暂无评论

Domain Adaptation in Visual Reinforcement Learning via Self-...

IEEE/RSJ International Conference on Intelligent robots and Systems (IROS)

作者： Lin Chen Jianan Huang Zhen Zhou Yaonan Wang Yang Mo Zhiqiang Miao Kai Zeng Mingtao Feng Danwei Wang School of Electrical and Information Engineering Hunan University Changsha China National Engineering Research Center for Robot Visual Perception and Control Technology Changsha China School of Computer Science and Technology Xidian University Xian China School of Electrical and Electrical Engineering Nanyang Technological University Singapore

ISBN: (数字)9798350377705

ISBN: (纸本)9798350377712

Generalizing visual reinforcement learning is fundamental to robot visual navigation, involving the acquisition of a policy from interactions with source environments to facilitate adaptation to analogous, yet unfamiliar target environments. Recent advancements capitalize on data augmentation techniques, self-supervised learning methods, and the generative adversarial network framework to train policy neural networks with enhanced generalizability. However, current methods, upon extracting domain-general latent features, further utilize these features to train the reinforcement learning policy, resulting in a decline in the performance of the learned policy guiding the agent to accomplish tasks. To tackle these challenges, a framework of self-expert imitation with purifying latent features was devised, empowering the policy to achieve robust and stable zero-shot generalization performance in visually similar domains previously unseen, without diminishing the performance of guiding the agent to accomplish tasks. The extraction method of domain-general latent features is proposed to enhance their quality based on the variational autoencoder. Extensive experiments have shown that our policy, compared with state-of-the-art counterparts, does not diminish the performance of the policy guiding the agent to accomplish tasks after generalization.

关键词： Training visualization Navigation Semantic segmentation Autoencoders Neural networks Reinforcement learning Self-supervised learning Feature extraction Intelligent robots

Soft Masked Transformer for Point Cloud Processing with Skip Attention-Based Upsampling

学校读者我要写书评

暂无评论

arXiv 2024年

作者： He, Yong Yu, Hongshan Ibrahim, Muhammad Liu, Xiaoyan Chen, Tongjia Ulhaq, Anwaar Mian, Ajmal The National Engineering Laboratory for Robot Visual Perception and Control Technology College of Electrical and Information Engineering Hunan University Lushan South Rd. Yuelu Dist. Changsha410082 China The Department of Computer Science The University of Western Australia WA6009 Australia Central Queensland University Sydney Campus Australia

Point cloud processing methods leverage local and global point features to cater to downstream tasks, yet they often overlook the task-level context inherent in point clouds during the encoding stage. We argue that integrating task-level information into the encoding stage significantly enhances performance. To that end, we propose SMTransformer which incorporates task-level information into a vector-based transformer by utilizing a soft mask generated from task-level queries and keys to learn the attention weights. Additionally, to facilitate effective communication between features from the encoding and decoding layers in high-level tasks such as segmentation, we introduce a skip-attention-based up-sampling block. This block dynamically fuses features from various resolution points across the encoding and decoding layers. To mitigate the increase in network parameters and training time resulting from the complexity of the aforementioned blocks, we propose a novel shared position encoding strategy. This strategy allows various transformer blocks to share the same position information over the same resolution points, thereby reducing network parameters and training time without compromising accuracy. Experimental comparisons with existing methods on multiple datasets demonstrate the efficacy of SMTransformer and skip-attention-based up-sampling for point cloud processing tasks, including semantic segmentation and classification. In particular, we achieve state-of-the-art semantic segmentation results of 73.4% mIoU on S3DIS Area 5 and 62.4% mIoU on SWAN dataset. Copyright © 2024, The Authors. All rights reserved.

关键词： Decoding

Full Point Encoding for Local Feature Aggregation in 3D Point Clouds

学校读者我要写书评

暂无评论

arXiv 2023年

作者： He, Yong Yu, Hongshan Yang, Zhengeng Liu, Xiaoyan Sun, Wei Mian, Ajmal National Engineering Laboratory for Robot Visual Perception and Control Technology College of Electrical and Information Engineering Hunan University Lushan South Rd. Yuelu Dist. Changsha410082 China Department of Computer Science The University of Western Australia WA6009 Australia

Point cloud processing methods exploit local point features and global context through aggregation which does not explicity model the internal correlations between local and global features. To address this problem, we propose full point encoding which is applicable to convolution and transformer architectures. Specifically, we propose Full Point Convolution (FPConv) and Full Point Transformer (FPTransformer) architectures. The key idea is to adaptively learn the weights from local and global geometric connections, where the connections are established through local and global correlation functions respectively. FPConv and FPTransformer simultaneously model the local and global geometric relationships as well as their internal correlations, demonstrating strong generalization ability and high performance. FPConv is incorporated in classical hierarchical network architectures to achieve local and global shape-aware learning. In FPTransformer, we introduce full point position encoding in self-attention, that hierarchically encodes each point position in the global and local receptive field. We also propose a shape aware downsampling block which takes into account the local shape and the global context. Experimental comparison to existing methods on benchmark datasets show the efficacy of FPConv and FPTransformer for semantic segmentation, object detection, classification, and normal estimation tasks. In particular, we achieve state-of-the-art semantic segmentation results of 76% mIoU on S3DIS 6-fold and 72.2% on S3DIS Area 5. Copyright © 2023, The Authors. All rights reserved.

关键词： Convolution

E-3DGS: Gaussian Splatting with Exposure and Motion Events

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Yin, Xiaoting Shi, Hao Bao, Yuhan Bing, Zhenshan Liao, Yiyi Yang, Kailun Wang, Kaiwei State Key Laboratory of Modern Optical Instrumentation Zhejiang University China School of Robotics National Engineering Research Center of Robot Visual Perception and Control Technology Hunan University China College of Information Science and Electronic Engineering Zhejiang University China The Chair of Robotics AI and Real-Time Systems Technical University of Munich Germany

Estimating Neural Radiance Fields (NeRFs) from images captured under optimal conditions has been extensively explored in the vision community. However, robotic applications often face challenges such as motion blur, insufficient illumination, and high computational overhead, which adversely affect downstream tasks like navigation, inspection, and scene visualization. To address these challenges, we propose E-3DGS, a novel event-based approach that partitions events into motion (from camera or object movement) and exposure (from camera exposure), using the former to handle fast-motion scenes and using the latter to reconstruct grayscale images for high-quality training and optimization of event-based 3D Gaussian Splatting (3DGS). We introduce a novel integration of 3DGS with exposure events for high-quality reconstruction of explicit scene representations. Our versatile framework can operate on motion events alone for 3D reconstruction, enhance quality using exposure events, or adopt a hybrid mode that balances quality and effectiveness by optimizing with initial exposure events followed by high-speed motion events. We also introduce EME-3D, a real-world 3D dataset with exposure events, motion events, camera calibration parameters, and sparse point clouds. Our method is faster and delivers better reconstruction quality than event-based NeRF while being more cost-effective than NeRF methods that combine event and RGB data by using a single event sensor. By combining motion and exposure events, E-3DGS sets a new benchmark for event-based 3D reconstruction with robust performance in challenging conditions and lower hardware demands. The source code and dataset will be available at https://***/MasterHow/E-3DGS. Copyright © 2024, The Authors. All rights reserved.

关键词： Gaussian distribution

Robust image-based landing control of a quadrotor on an unknown moving platform using circle features

学校读者我要写书评

暂无评论

Robust image-based landing control of a quadrotor on an unkn...

2021 IEEE International Conference on Real-Time Computing and robotics, RCAR 2021

作者： Lin, Jie Wang, Yaonan Miao, Zhiqiang Zhong, Hang Nie, Jingmou Fierro, Rafael College of Electrical and Information Engineering Hunan University Changsha410082 China National Engineering Laboratory for Robot Visual Perception and Control Changsha China University of New Mexico MARHES Lab Department of Electrical and Computer Engineering AlbuquerqueNM87131-0001 United States

ISBN: (纸本)9781665436786

This paper addresses the robust image-based landing control problem for a landing system, including an underactuated quadrotor and an unknown moving platform. Firstly, an image kinematics is constructed by using an image moment based on a circle feature in the virtual image plane, which decouples the landing system into a translation and a rotation control module. Exploiting the funnel control techniques with barrier Lyapunov functions (BLFs), the proposed IBVS control law does not depend on relative height needed by IBVS-like approaches and the explicit knowledge of moving platforms. Besides, it guarantees feature errors with prescribed transient and steady-state behavior. Finally, A simulation study is presented to verify the efficacy of the proposed control scheme. © 2021 IEEE.

关键词： Lyapunov functions

AdaptiveClick: Clicks-aware Transformer with Adaptive Focal Loss for Interactive Image Segmentation

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Lin, Jiacheng Chen, Jiajun Yang, Kailun Roitberg, Alina Li, Siyu Li, Zhiyong Li, Shutao The College of Computer Science and Electronic Engineering Hunan University Changsha410082 China The School of Robotics The National Engineering Research Center of Robot Visual Perception and Control Technology Hunan University Changsha410082 China The Institute for Artificial Intelligence The University of Stuttgart Stuttgart70569 Germany The College of Electrical and Information Engineering The Key Laboratory of Visual Perception and Artificial Intelligence of Hunan Province Hunan University Changsha410082 China

Interactive Image Segmentation (IIS) has emerged as a promising technique for decreasing annotation time. Substantial progress has been made in pre- and post-processing for IIS, but the critical issue of interaction ambiguity, notably hindering segmentation quality, has been under-researched. To address this, we introduce ADAPTIVECLICK – a click-aware transformer incorporating an adaptive focal loss that tackles annotation inconsistencies with tools for mask- and pixel-level ambiguity resolution. To the best of our knowledge, AdaptiveClick is the first transformer-based, mask-adaptive segmentation framework for IIS. The key ingredient of our method is the Click-Aware Mask-adaptive transformer Decoder (CAMD), which enhances the interaction between click and image features. Additionally, AdaptiveClick enables pixel-adaptive differentiation of hard and easy samples in the decision space, independent of their varying distributions. This is primarily achieved by optimizing a generalized Adaptive Focal Loss (AFL) with a theoretical guarantee, where two adaptive coefficients control the ratio of gradient values for hard and easy pixels. Our analysis reveals that the commonly used Focal and BCE losses can be considered special cases of the proposed AFL. With a plain ViT backbone, extensive experimental results on nine datasets demonstrate the superiority of AdaptiveClick compared to state-of-the-art methods. The source code is publicly available at https://***/lab206/AdaptiveClick. Copyright © 2023, The Authors. All rights reserved.

关键词： Pixels

PanoVPR: Towards Unified Perspective-to-Equirectangular visual Place Recognition via Sliding Windows across the Panoramic View

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Shi, Ze Shi, Hao Yang, Kailun Yin, Zhe Lin, Yining Wang, Kaiwei State Key Laboratory of Modern Optical Instrumentation Zhejiang University Hangzhou China School of Robotics and with the National Engineering Research Center of Robot Visual Perception and Control Technology Hunan University Changsha China Shanghai SUPREMIND Technology Co Ltd Shanghai China

— visual place recognition has gained significant attention in recent years as a crucial technology in autonomous driving and robotics. Currently, the two main approaches are the perspective view retrieval (P2P) paradigm and the equirectangular image retrieval (E2E) paradigm. However, it is practical and natural to assume that users only have consumer-grade pinhole cameras to obtain query perspective images and retrieve them in panoramic database images from map providers. To address this, we propose PanoVPR, a perspective-to-equirectangular (P2E) visual place recognition framework that employs sliding windows to eliminate feature truncation caused by hard cropping. Specifically, PanoVPR slides windows over the entire equirectangular image and computes feature descriptors for each window, which are then compared to determine place similarity. Notably, our unified framework enables direct transfer of the backbone from P2P methods without any modification, supporting not only CNNs but also Transformers. To facilitate training and evaluation, we derive the Pitts250k-P2E dataset from the Pitts250k and establish YQ360, latter is the first P2E visual place recognition dataset collected by a mobile robot platform aiming to simulate real-world task scenarios better. Extensive experiments demonstrate that PanoVPR achieves state-of-the-art performance and obtains 3.8% and 8.0% performance gain on Pitts250k-P2E and YQ360 compared to the previous best method, respectively. Code and datasets will be publicly available at PanoVPR. Copyright © 2023, The Authors. All rights reserved.

关键词： Query processing

HierDAMap: Towards Universal Domain Adaptive BEV Mapping via Hierarchical Perspective Priors

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Li, Siyu Cao, Yihong Shi, Hao Zang, Yongsheng He, Xuan Yang, Kailun Li, Zhiyong School of Robotics the National Engineering Research Center of Robot Visual Perception and Control Technology Hunan University Changsha410082 China State Key Laboratory of Extreme Photonics and Instrumentation Zhejiang University Hangzhou310027 China College of Computer Science and Electronic Engineering Hunan University Changsha410082 China Hunan Vanguard Group Corporation Limited Changsha410100 China

The exploration of Bird’s-Eye View (BEV) mapping technology has driven significant innovation in visual perception technology for autonomous driving. BEV mapping models need to be applied to the unlabeled real world, making the study of unsupervised domain adaptation models an essential path. However, research on unsupervised domain adaptation for BEV mapping remains limited and cannot perfectly accommodate all BEV mapping tasks. To address this gap, this paper proposes HierDAMap, a universal and holistic BEV domain adaptation framework with hierarchical perspective priors. Unlike existing research that solely focuses on image-level learning using prior knowledge, this paper explores the guiding role of perspective prior knowledge across three distinct levels: global, sparse, and instance levels. With these priors, HierDA consists of three essential components, including Semantic-Guided Pseudo Supervision (SGPS), Dynamic-Aware Coherence Learning (DACL), and Cross-Domain Frustum Mixing (CDFM). SGPS constrains the cross-domain consistency of perspective feature distribution through pseudo labels generated by vision foundation models in 2D space. To mitigate feature distribution discrepancies caused by spatial variations, DACL employs uncertainty-aware predicted depth as an intermediary to derive dynamic BEV labels from perspective pseudo-labels, thereby constraining the coarse BEV features derived from corresponding perspective features. CDFM, on the other hand, leverages perspective masks of view frustum to mix multi-view perspective images from both domains, which guides cross-domain view transformation and encoding learning through mixed BEV labels. Furthermore, this paper introduces intra-domain feature exchange data augmentation to enhance the efficiency of domain adaptation learning. The proposed method is verified on multiple BEV mapping tasks, such as BEV semantic segmentation, high-definition semantic, and vectorized mapping. It demonstrates competitive perfo

关键词： Birds