NTU RGB+D 60

The NTU-RGB+D 60 dataset is a benchmark that contains a collection of human action recognition data, specifically focusing on 60 action classes, and is used to evaluate models' performance in recognizing in-distribution actions and detecting out-of-distribution actions.

Papers using NTU RGB+D 60 (2)

Spatial Temporal Transformer Network For Skeleton-based Action Recognition2020 · 236 citesSpatial Temporal Transformer Network for Skeleton-based Action Recognition2020