Skeleton-based sign language recognition using a dual-stream spatio-temporal dynamic graph convolutional network