文献类型：期刊文献

英文题名：Efficient Detection Framework Adaptation for Edge Computing: A Plug-and-play Neural Network Toolbox Enabling Edge Deployment

作者：Wu, Jiaqi[1,2]; Zhang, Shihao[1]; Chen, Simin[3]; Wang, Lixu[4]; Wang, Zehua[6]; Chen, Wei[5]; He, Fangyuan[11]; Tian, Zijian[1]; Yu, F. Richard[7]; Leung, Victor C.M.[8,9,10]

第一作者：Wu, Jiaqi

机构：[1] School of Artificial Intelligence, China University of Mining and Technology [Beijing], Beijing, 100083, China; [2] Department of Electrical and Computer Engineering, University of British Columbia, Vancouver, 2332 Main Mall, Vancouver, BC, V6T 1Z4, Canada; [3] University of Texas at dallas, 800 W Campbell Rd, Richardson, TX, 75080, United States; [4] Northwestern University, 633 Clark St, Evanston, IL, 60208, United States; [5] School of Computer Science and Technology, China University of Mining and Technology, Jiangsu province, Xuzhou, 221116, China; [6] Department of Electrical and Computer Engineering, The University of British Columbia, Vancouver, 2332 Main Mall, Vancouver, BC, V6T 1Z4, Canada; [7] Department of Systems and Computer Engineering, Carleton University, Ottawa, ON, K1S 5B6, Canada; [8] Artificial Intelligence Research Institute, Shenzhen MSU-BIT University, Shenzhen, 518172, China; [9] College of Computer Science and Software Engineering, Shenzhen University, Shenzhen, 528060, China; [10] Department of Electrical and Computer Engineering, The University of British Columbia, Vancouver, BC, V6T 1Z4, Canada; [11] College of Applied Science and Technology, Beijing Union University, Beijing, 100012, China

第一机构：School of Artificial Intelligence, China University of Mining and Technology [Beijing], Beijing, 100083, China

年份：2024

外文期刊名：arXiv

收录：EI(收录号：20250014506)

语种：英文

外文关键词：Convolutional neural networks - Edge detection - Image annotation - Mobile edge computing - Security systems

摘要：Recently, edge computing has emerged as a prevailing paradigm in applying deep learning-based object detection models, offering a promising solution for time-sensitive tasks. However, existing edge object detection faces several challenges: 1) These methods struggle to balance detection precision and model lightweightness. 2) Existing generalized edge-deployment designs offer limited adaptability for object detection. 3) Current works lack real-world evaluation and validation. To address these challenges, we propose the Edge Detection Toolbox (ED-TOOLBOX), which leverages generalizable plug-and-play components to enable edge-site adaptation of object detection models. Specifically, we propose a lightweight Reparameterized Dynamic Convolutional Network (Rep-DConvNet) that employs a weighted multi-shape convolutional branch structure to enhance detection performance. Furthermore, ED-TOOLBOX includes a Sparse Cross-Attention (SC-A) network that adopts a localized-mapping-assisted self-attention mechanism to facilitate a well-crafted Joint Module in adaptively transferring features for further performance improvement. In real-world implementation and evaluation, we incorporate an Efficient Head into the popular You-Only-Look-Once (YOLO) approach to achieve faster edge model optimization. Moreover, we identify that helmet detection—one of the most representative edge object detection tasks—overlooks band fastening, which introduces potential safety hazards. To address this, we build a Helmet Band Detection Dataset (HBDD) and apply an edge object detection model optimized by the ED-TOOLBOX to tackle this real-world task. Extensive experiments validate the effectiveness of components in ED-TOOLBOX. In visual surveillance simulations, ED-TOOLBOX-assisted edge detection models outperform six state-of-the-art methods, enabling real-time and accurate detection. These results demonstrate that our approach offers a superior solution for edge object detection. Copyright ? 2024, The Authors. All rights reserved.