Two Stream Network for Stroke Detection in Table Tennis

TL;DR

This paper introduces a two-stream CNN approach for detecting strokes in table tennis videos, leveraging RGB and optical flow data, and participated in the MediaEval 2021 benchmark, achieving the best among competitors despite not surpassing the baseline.

Contribution

Proposes a two-stream CNN method for stroke detection in table tennis videos, integrating RGB and optical flow data, as part of a benchmark challenge.

Findings

Achieved the highest mAP among participants in the MediaEval 2021 Sport task.

Did not outperform the baseline on the test set.

Demonstrated the effectiveness of two-stream CNNs for sports video analysis.

Abstract

This paper presents a table tennis stroke detection method from videos. The method relies on a two-stream Convolutional Neural Network processing in parallel the RGB Stream and its computed optical flow. The method has been developed as part of the MediaEval 2021 benchmark for the Sport task. Our contribution did not outperform the provided baseline on the test set but has performed the best among the other participants with regard to the mAP metric.