【导读】本文整理了视频目标识别的数据集、论文、代码等,并对它们的性能进行了比较。
数据集
ImageNet VID Chanllenge
网站:
http://image-net.org/challenges/LSVRC/2017/#vid
Kaggle:
https://www.kaggle.com/account/login?returnUrl=%2Fc%2Fimagenet-object-detection-from-video-challenge
VisDrone Challenge
网站:
http://aiskyeye.com/
论文集
2016
Seq-NMS for video object detection
https://arxiv.org/abs/1602.08465
动机:在时间轴上,使预测的目标框更为光滑
总结:根据不同相邻时刻的预测构建时序图,然后使用动态规划选择整体最好的序列。
T-CNN: Tubelets with Convolutional Neural Networks for Object Detection from Videos
https://arxiv.org/abs/1604.02532
https://github.com/myfavouritekk/T-CNN
总结:使用跟踪算法,选择时序一致的高置信度的检测结果
Object Detection from Video Tubelets with Convolutional Neural Networks
https://arxiv.org/abs/1604.04053
https://github.com/myfavouritekk/vdetlib
Deep Feature Flow for Video Recognition
https://arxiv.org/abs/1611.07715
https://github.com/msracver/Deep-Feature-Flow
2017
Object Detection in Videos with Tubelet Proposal Networks
https://arxiv.org/abs/1702.06355
Flow-Guided Feature Aggregation for Video Object Detection
https://arxiv.org/abs/1703.10025
https://github.com/msracver/Flow-Guided-Feature-Aggregation
Detect to Track and Track to Detect
https://arxiv.org/abs/1710.03958
https://github.com/ZHANGHeng19931123/awesome-video-object-detection/blob/master/X.md
https://github.com/feichtenhofer/Detect-Track
Towards High Performance Video Object Detection
https://arxiv.org/abs/1711.11577
Video Object Detection with an Aligned Spatial-Temporal Memory
https://arxiv.org/abs/1712.06317
https://github.com/ZHANGHeng19931123/awesome-video-object-detection/blob/master/STMN.md
http://fanyix.cs.ucdavis.edu/project/stmn/project.html
https://www.youtube.com/watch?v=Vs3LqY1s9GY
2018
Object Detection in Videos by High Quality Object Linking
https://arxiv.org/abs/1801.09823
https://arxiv.org/abs/1804.05830
https://arxiv.org/abs/1804.05472、
https://github.com/ZHANGHeng19931123/awesome-video-object-detection/blob/master/X.md
https://github.com/hellock/scale-time-lattice
https://arxiv.org/abs/1803.05549
https://github.com/ZHANGHeng19931123/awesome-video-object-detection/blob/master/STSN.md
http://openaccess.thecvf.com/content_ECCV_2018/html/Shiyao_Wang_Fully_Motion-Aware_Network_ECCV_2018_paper.html
https://github.com/ZHANGHeng19931123/awesome-video-object-detection/blob/master/MANet.md
https://arxiv.org/abs/1811.11167
https://github.com/ZHANGHeng19931123/awesome-video-object-detection/blob/master/X.md
https://arxiv.org/pdf/1902.02910.pdf
https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=2&ved=2ahUKEwjWwNWa95_iAhUMyoUKHR-GAJwQFjABegQIBBAC&url=http%3A%2F%2Fwww.insticc.org%2FPrimoris%2FResources%2FPaperPdf.ashx%3FidPaper%3D72600&usg=AOvVaw1dTqzUoybpNRVkCdkA1xg0
https://github.com/ZHANGHeng19931123/seq_nms_yolo
性能比较:
Github链接:
https://github.com/ZHANGHeng19931123/awesome-video-object-detection
-END-
专 · 知
专知,专业可信的人工智能知识分发,让认知协作更快更好!欢迎登录www.zhuanzhi.ai,注册登录专知,获取更多AI知识资料!
欢迎微信扫一扫加入专知人工智能知识星球群,获取最新AI专业干货知识教程视频资料和与专家交流咨询!
请加专知小助手微信(扫一扫如下二维码添加),加入专知人工智能主题群,咨询技术商务合作~
专知《深度学习:算法到实战》课程全部完成!550+位同学在学习,现在报名,限时优惠!网易云课堂人工智能畅销榜首位!
点击“阅读原文”,了解报名专知《深度学习:算法到实战》课程