Challenge Results

Track 1 - FineAction Challenge on Temporal Action Localization

In this challenge, we have received 84 valid submissions for final evaluation. Winner teams’ performance and reports can be found below.

Rank Team Avg.mAP
1st Place Winners Beijing Jiaotong University & Institute of Automation [report] 21.97
2nd Place Winners Sun Yat-sen University [report] 16.49
3rd Place Winners OPPO Research Institute & Shanghai Jiao Tong University [report] 16.12

Track 2 - MultiSports Challenge on Human Action Detection

In this challenge, we have received 33 valid submissions for final evaluation. Winner teams’ performance and reports can be found below.

Rank Team V @ 0.1:0.9
1st Place Winners Computer Vision Lab, ETH Zurich [report] 31.709
2nd Place Winners Department of Computer Science, National Tsing Hua University, Taiwan [report] 12.843
3rd Place Winners Intelligent Media Analysis Group, Nanjing University of Science and Technology [report] 12.378

Track 3 - SportsMOT Challenge on Multi-actor Tracking

In this challenge, we have received 210 valid submissions for final evaluation. Winner teams’ performance and reports can be found below.

Rank Team HOTA
1st Place Winners BOE AIoT CTO [report] 76.264
2nd Place Winners School of Artificial Intelligence, Xidian University [report] 74.899
3rd Place Winners University of Washington, Seattle & Electronics and Telecommunications Research Institute, Korea [report] 73.968

Track 4 - Kinetics-TPS Challenge on Part-level Action Parsing

In this challenge, we have received 51 valid submissions for final evaluation. Winner teams’ performance and reports can be found below.

Rank Team Acc
1st Place Winners Shanghai Paidao Intelligent Technology Co., Ltd. & Chongqing Jiaotong University [report] 0.74
2nd Place Winners University of Electronic Science and Technology of China, Chengdu & Pengcheng Lab, Shenzhen [report] 0.69
3rd Place Winners University of Electronic Science and Technology of China, Chengdu [report] 0.66
4th Place Winners Sichuan University, Chengdu [report] 0.25

Track 5 - UrbanPipe Challenge on Fine-grained Video Anomaly Recognition

In this challenge, we have received 1393 valid submissions for final evaluation. Winner teams’ performance and reports can be found below.

Rank Team mAP
1st Place Winners Shanghai Paidao Intelligent Technology Co., Ltd. [report] 72.930
2nd Place Winners Key Laboratory of Intelligent Perception and Image Understanding,School of Artificial Intelligence Xidian University, Xi’an, China [report] 72.923
3rd Place Winners University of Science and Technology of China & Zhejiang University [report] 71.634