DeepEyes: Incentivizing "Thinking with Images" via Reinforcement Learning
This addresses the problem of multimodal reasoning for AI systems, offering a novel approach to enhance visual-text integration, though it appears incremental in advancing existing VLM capabilities.
The paper tackles the challenge of integrating visual and textual reasoning in Vision-Language Models by introducing DeepEyes, which uses reinforcement learning to incentivize 'thinking with images' and achieves significant performance gains on benchmarks for perception, reasoning, grounding, hallucination, and mathematical tasks.
Large Vision-Language Models (VLMs) have shown strong capabilities in multimodal understanding and reasoning, yet they are primarily constrained by text-based reasoning processes. However, achieving seamless integration of visual and textual reasoning which mirrors human cognitive processes remains a significant challenge. In particular, effectively incorporating advanced visual input processing into reasoning mechanisms is still an open question. Thus, in this paper, we explore the interleaved multimodal reasoning paradigm and introduce DeepEyes, a model with "thinking with images" capabilities incentivized through end-to-end reinforcement learning without the need for cold-start SFT. Notably, this ability emerges natively within the model itself, leveraging its inherent grounding ability as a tool instead of depending on separate specialized models. Specifically, we propose a tool-use-oriented data selection mechanism and a reward strategy to encourage successful tool-assisted reasoning trajectories. DeepEyes achieves significant performance gains on fine-grained perception and reasoning benchmarks and also demonstrates improvement in grounding, hallucination, and mathematical reasoning tasks. Interestingly, we observe the distinct evolution of tool-calling behavior from initial exploration to efficient and accurate exploitation, and diverse thinking patterns that closely mirror human visual reasoning processes. Code is available at https://github.com/Visual-Agent/DeepEyes.