Previous research on 3D skeleton-based human action recognition has frequently relied on a sequence-wise viewpoint normalization process. which adjusts the view directions of all segmented action sequences. This type of approach typically demonstrates robustness against variations in viewpoint found in short-term videos. a characteristic commonly encountered in public datasets. https://www.socialstudiesscholar.com/