Skip to Main content Skip to Navigation
Conference papers

Efficient Multi-stream Temporal Learning and Post-fusion Strategy for 3D Skeleton-based Hand Activity Recognition

Abstract : Recognizing first-person hand activity is a challenging task, especially when not enough data are available. In this paper, we tackle this challenge by proposing a new hybrid learning pipeline for skeleton-based hand activity recognition, which is composed of three blocks. First, for a given sequence of hand’s joint positions, the spatial features are extracted using a dedicated combination of local and global spatial hand-crafted features. Then, the temporal dependencies are learned using a multi-stream learning strategy. Finally, a hand activity sequence classifier is learned, via our Post-fusion strategy, applied to the previously learned temporal dependencies. The experiments, evaluated on two real-world data sets, show that our approach performs better than the state-of-the-art approaches. For more ablation studies, we compared our Post-fusion strategy with three traditional fusion baselines and showed an improvement above 2.4% of accuracy.
Complete list of metadata

https://hal.archives-ouvertes.fr/hal-03145521
Contributor : Admin HAL UR1 Connect in order to contact the contributor
Submitted on : Thursday, May 27, 2021 - 3:37:44 PM
Last modification on : Wednesday, April 27, 2022 - 4:27:03 AM
Long-term archiving on: : Saturday, August 28, 2021 - 6:07:49 PM

File

102327.pdf
Publisher files allowed on an open archive

Identifiers

Citation

Yasser Mohamed Boutaleb, Catherine Soladie, Nam-Duong Duong, Amine Kacete, Jérôme Royan, et al.. Efficient Multi-stream Temporal Learning and Post-fusion Strategy for 3D Skeleton-based Hand Activity Recognition. 16th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications (VISIGRAPP), Feb 2021, Online, France. pp.293-302, ⟨10.5220/0010232702930302⟩. ⟨hal-03145521⟩

Share

Metrics

Record views

178

Files downloads

93