A Closer Look at the Joint Training of Object Detection and Re-Identification in Multi-Object Tracking

判别式 计算机科学 目标检测 人工智能 对象(语法) 推论 假阳性悖论 光学(聚焦) 鉴定(生物学) 机器学习 任务(项目管理) 视频跟踪 特征(语言学) 基本事实 模式识别(心理学) 计算机视觉 工程类 植物 生物 系统工程 哲学 语言学 物理 光学
作者
Tianyi Liang,Baopu Li,Mengzhu Wang,Huibin Tan,Zhigang Luo
出处
期刊:IEEE transactions on image processing [Institute of Electrical and Electronics Engineers]
卷期号:32: 267-280 被引量:13
标识
DOI:10.1109/tip.2022.3227814
摘要

Unifying object detection and re-identification (ReID) into a single network enables faster multi-object tracking (MOT), while this multi-task setting poses challenges for training. In this work, we dissect the joint training of detection and ReID from two dimensions: label assignment and loss function. We find previous works generally overlook them and directly borrow the practices from object detection, inevitably causing inferior performance. Specifically, we identify a qualified label assignment for MOT should: 1) have the assignment cost aware of ReID cost, not just detection cost; 2) provide sufficient positive samples for robust feature learning while avoiding ambiguous positives (i.e., the positives shared by different ground-truth objects). To achieve the above goals, we first propose Identity-aware Label Assignment, which jointly considers the assignment cost of detection and ReID to select positive samples for each instance without ambiguities. Moreover, we advance a novel Discriminative Focal Loss that integrates ReID predictions with Focal Loss to focus the training on the discriminative samples. Finally, we upgrade the strong baseline FairMOT with our techniques and achieve up to 7.0 MOTA / 54.1% IDs improvements on MOT16/17/20 benchmarks under favorable inference speed, which verifies our tailored label assignment and loss function for MOT are superior to those inherited from object detection.
最长约 10秒,即可获得该文献文件

科研通智能强力驱动
Strongly Powered by AbleSci AI
更新
PDF的下载单位、IP信息已删除 (2025-6-4)

科研通是完全免费的文献互助平台,具备全网最快的应助速度,最高的求助完成率。 对每一个文献求助,科研通都将尽心尽力,给求助人一个满意的交代。
实时播报
Kiwi完成签到 ,获得积分10
1秒前
Weilang完成签到,获得积分10
1秒前
2秒前
SYX发布了新的文献求助10
2秒前
2秒前
CodeCraft应助酷酷珠采纳,获得10
3秒前
咕咚咕咚完成签到,获得积分10
3秒前
彳亍1117应助BREEZE采纳,获得10
3秒前
可爱的函函应助过氧化氢采纳,获得30
4秒前
yuyu完成签到,获得积分10
5秒前
plasma发布了新的文献求助10
5秒前
Orange应助ymlllym采纳,获得10
5秒前
6秒前
三土完成签到,获得积分10
6秒前
小俊发布了新的文献求助10
6秒前
耸耸完成签到 ,获得积分10
6秒前
gwd发布了新的文献求助10
7秒前
十二完成签到,获得积分10
7秒前
Akim应助44444采纳,获得10
7秒前
8秒前
8秒前
9秒前
9秒前
10秒前
Xide完成签到,获得积分10
10秒前
10秒前
11秒前
11秒前
Orange应助标致冬日采纳,获得10
11秒前
zhaohu47完成签到,获得积分10
12秒前
舒适的孤云完成签到,获得积分10
12秒前
12秒前
焦恩俊发布了新的文献求助10
15秒前
专注钢笔发布了新的文献求助10
15秒前
Jason发布了新的文献求助10
15秒前
华仔应助舒适的孤云采纳,获得10
15秒前
CR7应助耿耿采纳,获得20
16秒前
16秒前
所所应助赵怡宁采纳,获得10
16秒前
17秒前
高分求助中
The Mother of All Tableaux Order, Equivalence, and Geometry in the Large-scale Structure of Optimality Theory 2400
Ophthalmic Equipment Market by Devices(surgical: vitreorentinal,IOLs,OVDs,contact lens,RGP lens,backflush,diagnostic&monitoring:OCT,actorefractor,keratometer,tonometer,ophthalmoscpe,OVD), End User,Buying Criteria-Global Forecast to2029 2000
A new approach to the extrapolation of accelerated life test data 1000
Cognitive Neuroscience: The Biology of the Mind 1000
Cognitive Neuroscience: The Biology of the Mind (Sixth Edition) 1000
Optimal Transport: A Comprehensive Introduction to Modeling, Analysis, Simulation, Applications 800
Official Methods of Analysis of AOAC INTERNATIONAL 600
热门求助领域 (近24小时)
化学 材料科学 医学 生物 工程类 有机化学 生物化学 物理 内科学 纳米技术 计算机科学 化学工程 复合材料 遗传学 基因 物理化学 催化作用 冶金 细胞生物学 免疫学
热门帖子
关注 科研通微信公众号,转发送积分 3958850
求助须知:如何正确求助?哪些是违规求助? 3505102
关于积分的说明 11122496
捐赠科研通 3236558
什么是DOI,文献DOI怎么找? 1788899
邀请新用户注册赠送积分活动 871424
科研通“疑难数据库(出版商)”最低求助积分说明 802794