文献类型：会议论文

英文题名：SoccerNet 2022 Challenges Results

作者：Giancola, Silvio[1];Cioppa, Anthony[2];Deliege, Adrien[2];Magera, Floriane[2,3];Somers, Vladimir[4,5,6];Kang, Le[7];Zhou, Xin[7];Barnich, Olivier[8];De Vleeschouwer, Christophe[9];Alahi, Alexandre[10];Ghanem, Bernard[1];Van Droogenbroeck, Marc[2];Darwish, Abdulrahman[11];Maglo, Adrien[12];Clapes, Albert[13];Luyts, Andreas[14];Boiarov, Andrei[15];Xarles, Artur[16];Orcesi, Astrid[12];Shah, Avijit[17];Fan, Baoyu[18];Comandur, Bharath[19];Chen, Chen[20];Zhang, Chen[20,31];Zhao, Chen[21];Lin, Chengzhi[22];Chan, Cheuk-Yiu[23];Hui, Chun-Chuen[23];Li, Dengjie[24];Yang, Fan[25];Liang, Fan[24];Da, Fang[26];Yan, Feng[24];Yu, Fufu[27];Wang, Guanshuo[27];Chan, H. Anthony[23];Zhu, He[28];Kan, Hongwei[18];Chu, Jiaming[20,29];Hu, Jianming[28];Gu, Jianyang[20,30];Chen, Jin[31];Soares, Joao V. B.[17];Theiner, Jonas[32];De Corte, Jorge[33];Brito, Jose Henrique[34];Zhang, Jun[27];Li, Junjie[27,35];Liang, Junwei[27];Shen, Leqi[28];Ma, Lin[24];Chen, Lingchi[31];Marques, Miguel Santos[34];Azatov, Mike[36];Kasatkin, Nikita[15];Wang, Ning[20,37];Jia, Qiong[27];Pham, Quoc-Cuong[12];Ewerth, Ralph[32,38];Song, Ran[39];Li, Rengang[18];Gade, Rikke;Debien, Ruben;Zhang, Runze[18];Lee, Sangrok[40,41];Escalera, Sergio[13,16,42];Jiang, Shan[25];Odashima, Shigeyuki[25];Chen, Shimin[20];Masui, Shoichi[25];Ding, Shouhong;Chan, Sin-wai[23];Chen, Siyu[24];El-Shabrawy, Tallal[11];He, Tao[28];Moeslund, Thomas B.[13];Siu, Wan-Chi[23];Zhang, Wei[39];Li, Wei[20];Wang, Xiangwei[21];Tan, Xiao[43];Li, Xiaochuan[18];Wei, Xiaolin[24,31];Ye, Xiaoqing[43];Liu, Xing;Wang, Xinying[31];Guo, Yandong[20];Zhao, Yaqian[18];Yu, Yi[31];Li, Yingying[43];He, Yue[43];Zhong, Yujie[24];Guo, Zhenhua[18];Li, Zhiheng[39]

机构：[1]KAUST, Thuwal, Saudi Arabia;[2]Univ Liege, Liege, Belgium;[3]EVS Broadcast Equipment, Liege, Belgium;[4]Sportradar, London, England;[5]UCLouvain, London, England;[6]EPFL, London, England;[7]Baidu Res, Sunnyvale, CA USA;[8]EVS Broadcast Equipment, Liege, Belgium;[9]UCLouvain, Louvain La Neuve, Belgium;[10]Ecole Polytech Fed Lausanne, Lausanne, Switzerland;[11]German Univ Cairo, New Cairo, Egypt;[12]Univ Paris Saclay, CEA, List, Paris, France;[13]Aalborg Univ, Aalborg, Denmark;[14]ReBatch, Kontich, Belgium;[15]Schaffhausen Inst Technol, Schaffhausen, Switzerland;[16]Univ Barcelona, Barcelona, Spain;[17]Yahoo Res, Sunnyvale, CA USA;[18]Inspur Elect Informat Ind Co Ltd, State Key Lab Highend Server Storage Technol, Jinan, Peoples R China;[19]Purdue Univ, W Lafayette, IN USA;[20]OPPO Res Inst, Shenzhen, Peoples R China;[21]Baidu Inc, Dept Augmented Real Technol ART, Beijing, Peoples R China;[22]Sun Yat Sen Univ, Guangzhou, Peoples R China;[23]Caritas Inst Higher Educ Tseung Kwan O, Hong Kong, Peoples R China;[24]Meituan Inc, Beijing, Peoples R China;[25]Fujitsu Res, Kawasaki, Kanagawa, Japan;[26]QCraft Inc, Beijing, Peoples R China;[27]Tencent Youtu Lab, Shanghai, Peoples R China;[28]Tsinghua Univ, Beijing, Peoples R China;[29]Beijing Univ Posts & Telecommun Shenzhen, Beijing, Peoples R China;[30]Zhejiang Univ, Shenzhen, Peoples R China;[31]MGTV, Changsha, Peoples R China;[32]Leibniz Univ Hannover, Res Ctr L3S, Hannover, Germany;[33]ReBatch, Kontich, Belgium;[34]2Ai Sch Technol IPCA, Sao Martinho, Portugal;[35]Shanghai Jiao Tong Univ, Shanghai, Peoples R China;[36]Arsenal FC, London, England;[37]East China Univ Sci & Technol, Shenzhen, Peoples R China;[38]Leibniz Informat Ctr Sci & Technol, TIB, Hannover, Germany;[39]Shandong Univ, Sch Control Sci & Engn, Jinan, Peoples R China;[40]Yonsei Univ, Grad Sch Informat, Seoul, South Korea;[41]MODULABS, Seoul, South Korea;[42]Comp Vision Ctr, Barcelona, Spain; Shandong Univ, Jinan, Peoples R China;[43]Baidu Inc, Dept Comp Vision Technol VIS, Beijing, Peoples R China

会议论文集：5th ACM International Workshop on Multimedia Content Analysis in Sports (ACM MMSports)

会议日期：OCT 10-14, 2022

会议地点：Lisboa, PORTUGAL

语种：英文

外文关键词：datasets; challenges; computer vision; video understanding; neural networks; soccer

摘要：The SoccerNet 2022 challenges were the second annual video understanding challenges organized by the SoccerNet team. In 2022, the challenges were composed of 6 vision-based tasks: (1) action spotting, focusing on retrieving action timestamps in long untrimmed videos, (2) replay grounding, focusing on retrieving the live moment of an action shown in a replay, (3) pitch localization, focusing on detecting line and goal part elements, (4) camera calibration, dedicated to retrieving the intrinsic and extrinsic camera parameters, (5) player re-identi.cation, focusing on retrieving the same players across multiple views, and (6) multiple object tracking, focusing on tracking players and the ball through unedited video streams. Compared to last year's challenges, tasks (1-2) had their evaluation metrics rede.ned to consider tighter temporal accuracies, and tasks (3-6) were novel, including their underlying data and annotations. More information on the tasks, challenges and leaderboards are available on https://www.soccer-net.org. Baselines and development kits are available on https://github.com/SoccerNet.