Classification of Strokes in Table Tennis with a Three Stream Spatio-Temporal CNN for MediaEval 2020 - Archive ouverte HAL
Conference Papers Year : 2020

Classification of Strokes in Table Tennis with a Three Stream Spatio-Temporal CNN for MediaEval 2020

Abstract

This work presents a method for classifying table tennis strokes using spatio-temporal convolutional neural networks. The finegrained classification is performed on trimmed video segments recorded at 120 fps with different players performing in natural conditions. From those segments, the frames are extracted, their optical flow is computed and the pose of the player is estimated. From the optical flow amplitude, a region of interest is inferred. A three stream spatio-temporal convolutional neural network using combination of those modalities and 3D attention mechanisms is presented in order to perform classification.
Fichier principal
Vignette du fichier
MediaEval_2020_Working_Note_Paper.pdf (504.77 Ko) Télécharger le fichier
Origin Files produced by the author(s)

Dates and versions

hal-03104275 , version 1 (08-01-2021)

Identifiers

  • HAL Id : hal-03104275 , version 1

Cite

Pierre-Etienne Martin, Jenny Benois-Pineau, Boris Mansencal, Renaud Péteri, Julien Morlier. Classification of Strokes in Table Tennis with a Three Stream Spatio-Temporal CNN for MediaEval 2020. MediaEval 2020 Workshop, Dec 2020, Online, Unknown Region. ⟨hal-03104275⟩

Collections

CNRS
561 View
115 Download

Share

More