Semantic Softmax Loss for Zero-Shot Learning

VLM
Re-assign community