
Skeleton-based action recognition benefits from the use of vision transformers, providing robustness and efficiency in pseudo-image representation.


coremsg

skelvit-lightweight-skeleton-based-action-recognition-system-with-vision-transformers


SkelVIT: Lightweight Skeleton-Based Action Recognition System with Vision Transformers


title_rewrite


The author proposes SkelVIT, a three-level architecture utilizing vision transformers for skeleton-based action recognition. The study highlights the robustness of VITs on pseudo-image representation and the effectiveness of ensemble classifiers.


skelvit-vision-transformers-for-skeleton-based-action-recognition


SkelVIT: Vision Transformers for Skeleton-Based Action Recognition
