AI RESEARCH
EvoIR-Agent: Self-Evolving Image Restoration Agentic System via Experience-Driven Learning
arXiv CS.CV
•
ArXi:2605.22208v1 Announce Type: new Multimodal Large Language Model (MLLM)-driven image restoration agent nstrates effectiveness in degradation coupling scenarios by flexibly selecting tools and determining removal orders. However, their zero-shot planning often fails without experience, necessitating severe trial-and-error overhead to achieve satisfactory outcomes. Currently, two paradigms are employed to address this issue, yet a dilemma persists