提炼通道,实现高效深度跟踪

Shiming Ge, Zhao Luo, Chunhui Zhang, Yingying Hua, Dacheng Tao
{"title":"提炼通道,实现高效深度跟踪","authors":"Shiming Ge, Zhao Luo, Chunhui Zhang, Yingying Hua, Dacheng Tao","doi":"arxiv-2409.11785","DOIUrl":null,"url":null,"abstract":"Deep trackers have proven success in visual tracking. Typically, these\ntrackers employ optimally pre-trained deep networks to represent all diverse\nobjects with multi-channel features from some fixed layers. The deep networks\nemployed are usually trained to extract rich knowledge from massive data used\nin object classification and so they are capable to represent generic objects\nvery well. However, these networks are too complex to represent a specific\nmoving object, leading to poor generalization as well as high computational and\nmemory costs. This paper presents a novel and general framework termed channel\ndistillation to facilitate deep trackers. To validate the effectiveness of\nchannel distillation, we take discriminative correlation filter (DCF) and ECO\nfor example. We demonstrate that an integrated formulation can turn feature\ncompression, response map generation, and model update into a unified energy\nminimization problem to adaptively select informative feature channels that\nimprove the efficacy of tracking moving objects on the fly. Channel\ndistillation can accurately extract good channels, alleviating the influence of\nnoisy channels and generally reducing the number of channels, as well as\nadaptively generalizing to different channels and networks. The resulting deep\ntracker is accurate, fast, and has low memory requirements. Extensive\nexperimental evaluations on popular benchmarks clearly demonstrate the\neffectiveness and generalizability of our framework.","PeriodicalId":501130,"journal":{"name":"arXiv - CS - Computer Vision and Pattern Recognition","volume":null,"pages":null},"PeriodicalIF":0.0000,"publicationDate":"2024-09-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Distilling Channels for Efficient Deep Tracking\",\"authors\":\"Shiming Ge, Zhao Luo, Chunhui Zhang, Yingying Hua, Dacheng Tao\",\"doi\":\"arxiv-2409.11785\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Deep trackers have proven success in visual tracking. Typically, these\\ntrackers employ optimally pre-trained deep networks to represent all diverse\\nobjects with multi-channel features from some fixed layers. The deep networks\\nemployed are usually trained to extract rich knowledge from massive data used\\nin object classification and so they are capable to represent generic objects\\nvery well. However, these networks are too complex to represent a specific\\nmoving object, leading to poor generalization as well as high computational and\\nmemory costs. This paper presents a novel and general framework termed channel\\ndistillation to facilitate deep trackers. To validate the effectiveness of\\nchannel distillation, we take discriminative correlation filter (DCF) and ECO\\nfor example. We demonstrate that an integrated formulation can turn feature\\ncompression, response map generation, and model update into a unified energy\\nminimization problem to adaptively select informative feature channels that\\nimprove the efficacy of tracking moving objects on the fly. Channel\\ndistillation can accurately extract good channels, alleviating the influence of\\nnoisy channels and generally reducing the number of channels, as well as\\nadaptively generalizing to different channels and networks. The resulting deep\\ntracker is accurate, fast, and has low memory requirements. Extensive\\nexperimental evaluations on popular benchmarks clearly demonstrate the\\neffectiveness and generalizability of our framework.\",\"PeriodicalId\":501130,\"journal\":{\"name\":\"arXiv - CS - Computer Vision and Pattern Recognition\",\"volume\":null,\"pages\":null},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2024-09-18\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"arXiv - CS - Computer Vision and Pattern Recognition\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/arxiv-2409.11785\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"arXiv - CS - Computer Vision and Pattern Recognition","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/arxiv-2409.11785","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0

摘要

事实证明,深度跟踪器在视觉跟踪方面取得了成功。通常情况下,深度跟踪器采用经过优化预训练的深度网络,通过一些固定层的多通道特征来表示所有不同的物体。所使用的深度网络通常是为了从用于物体分类的海量数据中提取丰富的知识而训练的,因此它们能够很好地表示一般物体。然而,这些网络过于复杂,无法表示特定的运动物体,导致泛化能力差,计算和内存成本高。本文提出了一种新颖的通用框架,称为通道分散(channeldistillation),以促进深度跟踪器的发展。为了验证通道蒸馏的有效性,我们以判别相关滤波器(DCF)和 ECO 为例。我们证明,一个集成的公式可以将特征压缩、响应图生成和模型更新转化为一个统一的能量最小化问题,从而自适应地选择信息丰富的特征通道,提高对移动物体的实时跟踪效率。通道蒸馏可以准确地提取好的通道,减轻杂讯通道的影响,普遍减少通道数量,还能适应不同的通道和网络。由此产生的深度跟踪器准确、快速、内存要求低。在流行基准上进行的广泛实验评估清楚地证明了我们框架的有效性和通用性。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
Distilling Channels for Efficient Deep Tracking
Deep trackers have proven success in visual tracking. Typically, these trackers employ optimally pre-trained deep networks to represent all diverse objects with multi-channel features from some fixed layers. The deep networks employed are usually trained to extract rich knowledge from massive data used in object classification and so they are capable to represent generic objects very well. However, these networks are too complex to represent a specific moving object, leading to poor generalization as well as high computational and memory costs. This paper presents a novel and general framework termed channel distillation to facilitate deep trackers. To validate the effectiveness of channel distillation, we take discriminative correlation filter (DCF) and ECO for example. We demonstrate that an integrated formulation can turn feature compression, response map generation, and model update into a unified energy minimization problem to adaptively select informative feature channels that improve the efficacy of tracking moving objects on the fly. Channel distillation can accurately extract good channels, alleviating the influence of noisy channels and generally reducing the number of channels, as well as adaptively generalizing to different channels and networks. The resulting deep tracker is accurate, fast, and has low memory requirements. Extensive experimental evaluations on popular benchmarks clearly demonstrate the effectiveness and generalizability of our framework.
求助全文
通过发布文献求助,成功后即可免费获取论文全文。 去求助
来源期刊
自引率
0.00%
发文量
0
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
确定
请完成安全验证×
copy
已复制链接
快去分享给好友吧!
我知道了
右上角分享
点击右上角分享
0
联系我们:info@booksci.cn Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。 Copyright © 2023 布克学术 All rights reserved.
京ICP备2023020795号-1
ghs 京公网安备 11010802042870号
Book学术文献互助
Book学术文献互助群
群 号:481959085
Book学术官方微信