P2ANet: A Large-Scale Benchmark for Dense Action Detection from Table Tennis Match Broadcasting Videos

• Published in: ACM transactions on multimedia computing communications and applications Vol. 20; no. 4; pp. 1 - 23
• Main Authors: Bian, Jiang, Li, Xuhong, Wang, Tao, Wang, Qingzhong, Huang, Jun, Liu, Chen, Zhao, Jun, Lu, Feixiang, Dou, Dejing, Xiong, Haoyi
• Format: Journal Article
• Language: English
• Published: New York, NY ACM 01.04.2024
• Subjects: Activity recognition and understanding; Computing methodologies; Neural networks; Datasets; action recognition and localization; table tennis; annotation toolbox; video analysis
• ISSN: 1551-6857; 1551-6865
• DOI: 10.1145/3633516

While deep learning has been widely used for video analytics, such as video classification and action detection, dense action detection with fast-moving subjects from sports videos is still challenging. In this work, we release yet another sports video benchmark P2ANet for Ping Pong-Action detection, which consists of 2,721 video clips collected from the broadcasting videos of professional table tennis matches in World Table Tennis Championships and Olympiads. We work with a crew of table tennis professionals and referees on a specially designed annotation toolbox to obtain fine-grained action labels (in 14 classes) for every ping-pong action that appeared in the dataset, and formulate two sets of action detection problems—action localization and action recognition. We evaluate a number of commonly seen action recognition (e.g., TSM, TSN, Video SwinTransformer, and Slowfast) and action localization models (e.g., BSN, BSN++, BMN, TCANet), using P2ANet for both problems, under various settings. These models can only achieve 48% area under the AR-AN curve for localization and 82% top-one accuracy for recognition since the ping-pong actions are dense with fast-moving subjects but broadcasting videos are with only 25 FPS. The results confirm that P2ANet is still a challenging task and can be used as a special benchmark for dense action detection from videos. We invite readers to examine our dataset by visiting the following link: https://github.com/Fred1991/P2ANET.