Deep Affinity Network for Multiple Object Tracking

计算机科学 人工智能 视频跟踪 计算机视觉 深度学习 目标检测 帧(网络) 标杆管理 联想(心理学) 对象(语法) 模式识别(心理学) 跟踪(教育) 利用 哲学 营销 业务 心理学 认识论 电信 计算机安全 教育学
作者
Shijie Sun,Naveed Akhtar,Huansheng Song,Ajmal Mian,Mubarak Shah
出处
期刊:IEEE Transactions on Pattern Analysis and Machine Intelligence [Institute of Electrical and Electronics Engineers]
卷期号:: 1-1 被引量:143
标识
DOI:10.1109/tpami.2019.2929520
摘要

Multiple Object Tracking (MOT) plays an important role in solving many fundamental problems in video analysis and computer vision. Most MOT methods employ two steps: Object Detection and Data Association. The first step detects objects of interest in every frame of a video, and the second establishes correspondence between the detected objects in different frames to obtain their tracks. Object detection has made tremendous progress in the last few years due to deep learning. However, data association for tracking still relies on hand crafted constraints such as appearance, motion, spatial proximity, grouping etc. to compute affinities between the objects in different frames. In this paper, we harness the power of deep learning for data association in tracking by jointly modeling object appearances and their affinities between different frames in an end-to-end fashion. The proposed Deep Affinity Network (DAN) learns compact, yet comprehensive features of pre-detected objects at several levels of abstraction, and performs exhaustive pairing permutations of those features in any two frames to infer object affinities. DAN also accounts for multiple objects appearing and disappearing between video frames. We exploit the resulting efficient affinity computations to associate objects in the current frame deep into the previous frames for reliable on-line tracking. Our technique is evaluated on popular multiple object tracking challenges MOT15, MOT17 and UA-DETRAC. Comprehensive benchmarking under twelve evaluation metrics demonstrates that our approach is among the best performing techniques on the leader board for these challenges. The open source implementation of our work is available at https://github.com/shijieS/SST.git.
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
大幅提高文件上传限制,最高150M (2024-4-1)

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
雪碧呀完成签到 ,获得积分10
1秒前
公孙世往完成签到,获得积分10
1秒前
stop here发布了新的文献求助10
1秒前
Orange应助JXY采纳,获得10
1秒前
棒子面糊糊完成签到,获得积分10
1秒前
Mrwang完成签到,获得积分10
2秒前
2秒前
you完成签到,获得积分10
2秒前
CXSCXD完成签到,获得积分10
4秒前
着急的翠彤完成签到,获得积分10
4秒前
Lin.隽发布了新的文献求助10
5秒前
5秒前
5秒前
外向的妙旋完成签到 ,获得积分10
6秒前
文艺的青旋完成签到 ,获得积分10
6秒前
Akim应助wanci采纳,获得30
6秒前
7秒前
完美世界应助江江采纳,获得10
7秒前
花某人完成签到,获得积分20
8秒前
Owen应助xx采纳,获得10
8秒前
渣渣慧完成签到,获得积分10
9秒前
10秒前
11秒前
11秒前
伶俐一曲完成签到,获得积分10
11秒前
北极光完成签到,获得积分20
11秒前
11秒前
Sulin完成签到 ,获得积分10
12秒前
ding应助Lin.隽采纳,获得10
12秒前
李爱国应助三杠采纳,获得10
12秒前
12秒前
啵叽一口发布了新的文献求助10
13秒前
ss完成签到,获得积分10
13秒前
伶俐一曲发布了新的文献求助10
13秒前
薰硝壤应助StevenXiong采纳,获得10
13秒前
Bernard完成签到,获得积分10
14秒前
Candy完成签到,获得积分10
14秒前
14秒前
三七037应助感动冰姬采纳,获得10
14秒前
知性的友易完成签到,获得积分10
14秒前
高分求助中
Sustainability in Tides Chemistry 2800
The Young builders of New china : the visit of the delegation of the WFDY to the Chinese People's Republic 1000
Rechtsphilosophie 1000
Bayesian Models of Cognition:Reverse Engineering the Mind 888
Le dégorgement réflexe des Acridiens 800
Defense against predation 800
Very-high-order BVD Schemes Using β-variable THINC Method 568
热门求助领域 (近24小时)
化学 医学 生物 材料科学 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 基因 遗传学 催化作用 物理化学 免疫学 量子力学 细胞生物学
热门帖子
关注 科研通微信公众号,转发送积分 3135300
求助须知:如何正确求助?哪些是违规求助? 2786282
关于积分的说明 7776733
捐赠科研通 2442250
什么是DOI,文献DOI怎么找? 1298501
科研通“疑难数据库(出版商)”最低求助积分说明 625124
版权声明 600847