Tracking

DLT

Learning A Deep Compact Image Representation for Visual Tracking

• intro: NIPS 2013
• project page: http://winsty.net/dlt.html

Hierarchical Convolutional Features for Visual Tracking

• intro: ICCV 2015
• project page: https://sites.google.com/site/jbhuang0604/publications/cf2
• github: https://github.com/jbhuang0604/CF2

Robust Visual Tracking via Convolutional Networks

• arxiv: http://arxiv.org/abs/1501.04505
• paper: http://kaihuazhang.net/CNT.pdf
• code: http://kaihuazhang.net/CNT_matlab.rar

SO-DLT

Transferring Rich Feature Hierarchies for Robust Visual Tracking

• arxiv: http://arxiv.org/abs/1501.04587
• slides: http://valse.mmcheng.net/ftp/20150325/RVT.pptx

MDNet

Learning Multi-Domain Convolutional Neural Networks for Visual Tracking

• intro: The Winner of The VOT2015 Challenge
• keywords: Multi-Domain Network (MDNet)
• homepage: http://cvlab.postech.ac.kr/research/mdnet/
• arxiv: http://arxiv.org/abs/1510.07945
• github: https://github.com/HyeonseobNam/MDNet

RATM: Recurrent Attentive Tracking Model

• arxiv: http://arxiv.org/abs/1510.08660
• github: https://github.com/saebrahimi/RATM

Understanding and Diagnosing Visual Tracking Systems

• intro: ICCV 2015
• project page: http://winsty.net/tracker_diagnose.html
• paper: http://winsty.net/papers/diagnose.pdf
• code(Matlab): http://120.52.72.43/winsty.net/c3pr90ntcsf0/diagnose/diagnose_code.zip

Recurrently Target-Attending Tracking

• paper: http://www.cv-foundation.org/openaccess/content_cvpr_2016/html/Cui_Recurrently_Target-Attending_Tracking_CVPR_2016_paper.html
• paper: http://www.cv-foundation.org/openaccess/content_cvpr_2016/papers/Cui_Recurrently_Target-Attending_Tracking_CVPR_2016_paper.pdf

Visual Tracking with Fully Convolutional Networks

• intro: ICCV 2015
• paper: http://202.118.75.4/lu/Paper/ICCV2015/iccv15_lijun.pdf
• github: https://github.com/scott89/FCNT

Deep Tracking: Seeing Beyond Seeing Using Recurrent Neural Networks

• intro: AAAI 2016
• arxiv: http://arxiv.org/abs/1602.00991
• github: https://github.com/pondruska/DeepTracking

Learning to Track at 100 FPS with Deep Regression Networks

• intro: ECCV 2015
• intro: GOTURN: Generic Object Tracking Using Regression Networks
• project page: http://davheld.github.io/GOTURN/GOTURN.html
• arxiv: http://arxiv.org/abs/1604.01802
• github: https://github.com/davheld/GOTURN

Learning by tracking: Siamese CNN for robust target association

• arxiv: http://arxiv.org/abs/1604.07866

Fully-Convolutional Siamese Networks for Object Tracking

• intro: ECCV 2016
• intro: State-of-the-art performance in arbitrary object tracking at 50-100 FPS with Fully Convolutional Siamese networks
• project page: http://www.robots.ox.ac.uk/~luca/siamese-fc.html
• arxiv: http://arxiv.org/abs/1606.09549
• github(official): https://github.com/bertinetto/siamese-fc
• github(official): https://github.com/torrvision/siamfc-tf
• valse-video: http://www.iqiyi.com/w_19ruirwrel.html#vfrm=8-8-0-1

Hedged Deep Tracking

• project page(paper+code): https://sites.google.com/site/yuankiqi/hdt
• paper: https://docs.google.com/viewer?a=v&pid=sites&srcid=ZGVmYXVsdGRvbWFpbnx5dWFua2lxaXxneDoxZjc2MmYwZGIzNjFhYTRl

ROLO

Spatially Supervised Recurrent Convolutional Neural Networks for Visual Object Tracking

• intro: ROLO is short for Recurrent YOLO, aimed at simultaneous object detection and tracking
• project page: http://guanghan.info/projects/ROLO/
• arxiv: http://arxiv.org/abs/1607.05781
• github: https://github.com/Guanghan/ROLO

Visual Tracking via Shallow and Deep Collaborative Model

• arxiv: http://arxiv.org/abs/1607.08040

Beyond Correlation Filters: Learning Continuous Convolution Operators for Visual Tracking

• intro: ECCV 2016
• intro: OTB-2015 (+5.1% in mean OP), Temple-Color (+4.6% in mean OP), and VOT2015 (20% relative reduction in failure rate)
• keywords: Continuous Convolution Operator Tracker (C-COT)
• project page: http://www.cvl.isy.liu.se/research/objrec/visualtracking/conttrack/index.html
• arxiv: http://arxiv.org/abs/1608.03773
• github(MATLAB): https://github.com/martin-danelljan/Continuous-ConvOp

Predictive Vision Model (PVM)

Unsupervised Learning from Continuous Video in a Scalable Predictive Recurrent Network

• arxiv: http://arxiv.org/abs/1607.06854
• github: https://github.com/braincorp/PVM

Modeling and Propagating CNNs in a Tree Structure for Visual Tracking

• arxiv: http://arxiv.org/abs/1608.07242

Robust Scale Adaptive Kernel Correlation Filter Tracker With Hierarchical Convolutional Features

• paper: http://ieeexplore.ieee.org/document/7496863/

Deep Tracking on the Move: Learning to Track the World from a Moving Vehicle using Recurrent Neural Networks

• arxiv: https://arxiv.org/abs/1609.09365

OTB Results: visual tracker benchmark results

• github: https://github.com/foolwood/benchmark_results

Convolutional Regression for Visual Tracking

• arxiv: https://arxiv.org/abs/1611.04215

Semantic tracking: Single-target tracking with inter-supervised convolutional networks

• arxiv: https://arxiv.org/abs/1611.06395

SANet: Structure-Aware Network for Visual Tracking

• arxiv: https://arxiv.org/abs/1611.06878

ECO: Efficient Convolution Operators for Tracking

• intro: CVPR 2017
• project page: http://www.cvl.isy.liu.se/research/objrec/visualtracking/ecotrack/index.html
• arxiv: https://arxiv.org/abs/1611.09224
• github: https://github.com/martin-danelljan/ECO

Dual Deep Network for Visual Tracking

• arxiv: https://arxiv.org/abs/1612.06053

Deep Motion Features for Visual Tracking

• intro: ICPR 2016. Best paper award in the “Computer Vision and Robot Vision” track
• arxiv: https://arxiv.org/abs/1612.06615

Globally Optimal Object Tracking with Fully Convolutional Networks

• arxiv: https://arxiv.org/abs/1612.08274

Robust and Real-time Deep Tracking Via Multi-Scale Domain Adaptation

• arxiv: https://arxiv.org/abs/1701.00561
• bitbucket: https://bitbucket.org/xinke_wang/msdat

Tracking The Untrackable: Learning To Track Multiple Cues with Long-Term Dependencies

• arxiv: https://arxiv.org/abs/1701.01909

Large Margin Object Tracking with Circulant Feature Maps

• intro: CVPR 2017
• intro: The experimental results demonstrate that the proposed tracker performs superiorly against several state-of-the-art algorithms on the challenging benchmark sequences while runs at speed in excess of 80 frames per secon
• arxiv: https://arxiv.org/abs/1703.05020
• notes: https://zhuanlan.zhihu.com/p/25761718

DCFNet: Discriminant Correlation Filters Network for Visual Tracking

• arxiv: https://arxiv.org/abs/1704.04057
• github: https://github.com/foolwood/DCFNet

End-to-end representation learning for Correlation Filter based tracking

• intro: CVPR 2017. University of Oxford
• intro: Training a Correlation Filter end-to-end allows lightweight networks of 2 layers (600 kB) to achieve state-of-the-art performance in tracking, at high-speed.
• project page: http://www.robots.ox.ac.uk/~luca/cfnet.html
• arxiv: https://arxiv.org/abs/1704.06036
• gtihub: https://github.com/bertinetto/cfnet

Context-Aware Correlation Filter Tracking

• intro: CVPR 2017 Oral
• project page: https://ivul.kaust.edu.sa/Pages/pub-ca-cf-tracking.aspx
• paper: https://ivul.kaust.edu.sa/Documents/Publications/2017/Context-Aware%20Correlation%20Filter%20Tracking.pdf
• github: https://github.com/thias15/Context-Aware-CF-Tracking

Robust Multi-view Pedestrian Tracking Using Neural Networks

https://arxiv.org/abs/1704.06370

Re3 : Real-Time Recurrent Regression Networks for Object Tracking

• intro: University of Washington
• arxiv: https://arxiv.org/abs/1705.06368
• demo: https://www.youtube.com/watch?v=PC0txGaYz2I

Robust Tracking Using Region Proposal Networks

https://arxiv.org/abs/1705.10447

Hierarchical Attentive Recurrent Tracking

• intro: NIPS 2017. University of Oxford
• arxiv: https://arxiv.org/abs/1706.09262
• github: https://github.com/akosiorek/hart
• results: https://youtu.be/Vvkjm0FRGSs

Siamese Learning Visual Tracking: A Survey

https://arxiv.org/abs/1707.00569

Robust Visual Tracking via Hierarchical Convolutional Features

• project page: https://sites.google.com/site/chaoma99/hcft-tracking
• arxiv: https://arxiv.org/abs/1707.03816
• github: https://github.com/chaoma99/HCFTstar

CREST: Convolutional Residual Learning for Visual Tracking

• intro: ICCV 2017
• project page: http://www.cs.cityu.edu.hk/~yibisong/iccv17/index.html
• arxiv: https://arxiv.org/abs/1708.00225
• github: https://github.com/ybsong00/CREST-Release

Learning Policies for Adaptive Tracking with Deep Feature Cascades

• intro: ICCV 2017 Spotlight
• arxiv: https://arxiv.org/abs/1708.02973

Recurrent Filter Learning for Visual Tracking

• intro: ICCV 2017 Workshop on VOT
• arxiv: https://arxiv.org/abs/1708.03874

Correlation Filters with Weighted Convolution Responses

• intro: ICCV 2017 workshop. 5th visual object tracking(VOT) tracker CFWCR
• paper: http://openaccess.thecvf.com/content_ICCV_2017_workshops/papers/w28/He_Correlation_Filters_With_ICCV_2017_paper.pdf
• github: https://github.com/he010103/CFWCR

Semantic Texture for Robust Dense Tracking

https://arxiv.org/abs/1708.08844

Learning Multi-frame Visual Representation for Joint Detection and Tracking of Small Objects

https://arxiv.org/abs/1709.04666

Tracking Persons-of-Interest via Unsupervised Representation Adaptation

• intro: Northwestern Polytechnical University & Virginia Tech & Hanyang University
• keywords: Multi-face tracking
• project page: http://vllab1.ucmerced.edu/~szhang/FaceTracking/
• arxiv: https://arxiv.org/abs/1710.02139

End-to-end Flow Correlation Tracking with Spatial-temporal Attention

https://arxiv.org/abs/1711.01124

UCT: Learning Unified Convolutional Networks for Real-time Visual Tracking

• intro: ICCV 2017 Workshops
• arxiv: https://arxiv.org/abs/1711.04661

Pixel-wise object tracking

https://arxiv.org/abs/1711.07377

MAVOT: Memory-Augmented Video Object Tracking

https://arxiv.org/abs/1711.09414

Learning Hierarchical Features for Visual Object Tracking with Recursive Neural Networks

https://arxiv.org/abs/1801.02021

Parallel Tracking and Verifying

https://arxiv.org/abs/1801.10496

Saliency-Enhanced Robust Visual Tracking

https://arxiv.org/abs/1802.02783

A Twofold Siamese Network for Real-Time Object Tracking

• intro: CVPR 2018
• arxiv: https://arxiv.org/abs/1802.08817

Learning Dynamic Memory Networks for Object Tracking

https://arxiv.org/abs/1803.07268

Context-aware Deep Feature Compression for High-speed Visual Tracking

• intro: CVPR 2018
• arxiv: https://arxiv.org/abs/1803.10537

VITAL: VIsual Tracking via Adversarial Learning

• intro: CVPR 2018 Spotlight
• arixv: https://arxiv.org/abs/1804.04273

Unveiling the Power of Deep Tracking

https://arxiv.org/abs/1804.06833

A Novel Low-cost FPGA-based Real-time Object Tracking System

• intro: ASICON 2017
• arxiv: https://arxiv.org/abs/1804.05535

MV-YOLO: Motion Vector-aided Tracking by Semantic Object Detection

https://arxiv.org/abs/1805.00107

Multi-Object Tracking (MOT)

Virtual Worlds as Proxy for Multi-Object Tracking Analysis

• arxiv: http://arxiv.org/abs/1605.06457
• dataset(Virtual KITTI): http://www.xrce.xerox.com/Research-Development/Computer-Vision/Proxy-Virtual-Worlds

Multi-Person Tracking by Multicut and Deep Matching

• arxiv: http://arxiv.org/abs/1608.05404

Multi-Class Multi-Object Tracking using Changing Point Detection

• intro: changing point detection, entity transition, object detection from video, convolutional neural network
• arxiv: http://arxiv.org/abs/1608.08434

POI: Multiple Object Tracking with High Performance Detection and Appearance Feature

• intro: ECCV workshop BMTT 2016. Sensetime
• keywords: KDNT
• arxiv: https://arxiv.org/abs/1610.06136

Simple Online and Realtime Tracking with a Deep Association Metric

• arxiv: https://arxiv.org/abs/1703.07402
• mot challenge: https://motchallenge.net/tracker/DeepSORT_2
• github(official, Python): https://github.com/nwojke/deep_sort
• github(C++): https://github.com/oylz/ds

Deep Network Flow for Multi-Object Tracking

• intro: CVPR 2017
• arxiv: https://arxiv.org/abs/1706.08482

Online Multi-Object Tracking Using CNN-based Single Object Tracker with Spatial-Temporal Attention Mechanism

https://arxiv.org/abs/1708.02843

Recurrent Autoregressive Networks for Online Multi-Object Tracking

https://arxiv.org/abs/1711.02741

SOT for MOT

• intro: Tsinghua University & Megvii Inc. (Face++)
• arxiv: https://arxiv.org/abs/1712.01059

Multi-Target, Multi-Camera Tracking by Hierarchical Clustering: Recent Progress on DukeMTMC Project

https://arxiv.org/abs/1712.09531

Multiple Target Tracking by Learning Feature Representation and Distance Metric Jointly

https://arxiv.org/abs/1802.03252

Tracking Noisy Targets: A Review of Recent Object Tracking Approaches

https://arxiv.org/abs/1802.03098

Machine Learning Methods for Solving Assignment Problems in Multi-Target Tracking

• intro: University of Florida
• arxiv: https://arxiv.org/abs/1802.06897

Tracking by Prediction: A Deep Generative Model for Mutli-Person localisation and Tracking

• intro: WACV 2018. Queensland University of Technology (QUT)
• arxiv: https://arxiv.org/abs/1803.03347

Learning to Detect and Track Visible and Occluded Body Joints in a Virtual World

• intro: University of Modena and Reggio Emilia
• arxiv: https://arxiv.org/abs/1803.08319

Features for Multi-Target Multi-Camera Tracking and Re-Identification

• intro: CVPR 2018
• intro: https://arxiv.org/abs/1803.10859

Trajectory Factory: Tracklet Cleaving and Re-connection by Deep Siamese Bi-GRU for Multiple Object Tracking

• intro: Peking University
• arxiv: https://arxiv.org/abs/1804.04555

Tracking with Reinforcement Learning

Deep Reinforcement Learning for Visual Object Tracking in Videos

• intro: University of California at Santa Barbara & Samsung Research America
• arxiv: https://arxiv.org/abs/1701.08936

Visual Tracking by Reinforced Decision Making

• arxiv: https://arxiv.org/abs/1702.06291

End-to-end Active Object Tracking via Reinforcement Learning

https://arxiv.org/abs/1705.10561

Action-Decision Networks for Visual Tracking with Deep Reinforcement Learning

• project page: https://sites.google.com/view/cvpr2017-adnet
• paper: https://drive.google.com/file/d/0B34VXh5mZ22cZUs2Umc1cjlBMFU/view?usp=drive_web

Tracking as Online Decision-Making: Learning a Policy from Streaming Videos with Reinforcement Learning

https://arxiv.org/abs/1707.04991

Detect to Track and Track to Detect

• intro: ICCV 2017
• project page: https://www.robots.ox.ac.uk/~vgg/research/detect-track/
• arxiv: https://arxiv.org/abs/1710.03958
• github: https://github.com/feichtenhofer/Detect-Track

Projects

Tensorflow_Object_Tracking_Video

• intro: Object Tracking in Tensorflow ( Localization Detection Classification ) developed to partecipate to ImageNET VID competition
• github: https://github.com/DrewNF/Tensorflow_Object_Tracking_Video