Multimodal Generation Comparison Hub
3 papers - avg viability 7.0
Top Papers
- MER-Bench: A Comprehensive Benchmark for Multimodal Meme Reappraisal(8.0)
MER-Bench enables the transformation of negative memes into constructive ones through emotion-controllable multimodal generation.
- Learning to Generate via Understanding: Understanding-Driven Intrinsic Rewarding for Unified Multimodal Models(7.0)
Improve multimodal model generation quality by using the model's understanding branch to guide the generation process through a self-supervised reinforcement learning framework.
- Towards Unified Multimodal Interleaved Generation via Group Relative Policy Optimization(6.0)
A reinforcement learning strategy to enhance multimodal interleaved generation in existing unified vision-language models.