Awesome-Skeleton-based-Action-Recognition

Why we included this project

Getting into skeleton-based action recognition is easier with this list around. It collects the field's key papers and datasets, tracks leaderboards, and organizes everything by year and by learning paradigm, so you can follow how supervised methods evolved and see where semi-supervised and unsupervised work stands now. The entries also point to the standard NTU RGB+D benchmarks and to related repositories and portals that carry code and pretrained models, which saves real time when you want to reproduce a result. Because it is a curated index rather than runnable software, its value is orientation: it lets you survey the literature quickly and decide which papers and datasets deserve a closer look before you commit to an approach.