Multi-Task Video Captioning with Video and Entailment Generation

40001

Knowledge Triples

Source Entity Relationship Target Entity
improvedBy