Addressing target shift in zero-shot learning using grouped adversarial learning

VLM
Re-assign community