From Perception to Punchline: Empowering VLM with the Art of In-the-wild Meme

Xueyan Li, Yingyi Xue, Mengjie Jiang +2

Generating humorous memes is a challenging multimodal task that moves beyond direct image-to-caption supervision. It requires a nuanced reasoning over visual content, contextual cu…