Language-aware Visual Semantic Distillation for Video Question Answering | Read Paper on Bytez