-
MASA: Motion-aware Masked Autoencoder with Semantic Alignment for Sign Langua...
Sign language recognition (SLR) has long been plagued by insufficient model representation capabilities. Al-though current pre-training approaches have alleviated this dilemma...
