Fine-Grained Action Detection with RGB and Pose Information using Two Stream Convolutional Networks - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2023

Fine-Grained Action Detection with RGB and Pose Information using Two Stream Convolutional Networks

Résumé

As participants of the MediaEval 2022 Sport Task, we propose a two-stream network approach for the classification and detection of table tennis strokes. Each stream is a succession of 3D Convolutional Neural Network (CNN) blocks using attention mechanisms. Each stream processes different 4D inputs. Our method utilizes raw RGB data and pose information computed from MMPose toolbox. The pose information is treated as an image by applying the pose either on a black background or on the original RGB frame it has been computed from. Best performance is obtained by feeding raw RGB data to one stream, Pose + RGB (PRGB) information to the other stream and applying late fusion on the features. The approaches were evaluated on the provided TTStroke-21 data sets. We can report an improvement in stroke classification, reaching 87.3% of accuracy, while the detection does not outperform the baseline but still reaches an IoU of 0.349 and mAP of 0.110.
Fichier principal
Vignette du fichier
paper3423.pdf (1.75 Mo) Télécharger le fichier
main.pdf (1.74 Mo) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-03958848 , version 1 (30-01-2023)

Identifiants

  • HAL Id : hal-03958848 , version 1

Citer

Leonard Hacker, Finn Bartels, Pierre-Etienne Martin. Fine-Grained Action Detection with RGB and Pose Information using Two Stream Convolutional Networks. MediaEval 2022 Workshop, Jan 2023, Bergen, Norway. ⟨hal-03958848⟩
8 Consultations
18 Téléchargements

Partager

Gmail Facebook X LinkedIn More