CODE: Contrasting Self-generated Description to Combat Hallucination in Large Multi-modal Models

Open in new window