VisionReasoner: Unified Reasoning-Integrated Visual Perception via Reinforcement Learning

TL;DR

VisionReasoner is a unified reinforcement learning framework that enhances visual perception tasks by integrating reasoning, achieving superior performance across detection, segmentation, and counting tasks without relying on annotated reasoning data.

Contribution

The paper introduces VisionReasoner, a novel unified model that combines reasoning with visual perception tasks using reinforcement learning and a shared reward mechanism.

Findings

Outperforms baseline Qwen2.5VL on multiple perception tasks

Achieves 29.1% improvement on COCO detection

Demonstrates faithful and reliable reasoning without annotated data

Abstract

Large vision-language models exhibit inherent capabilities to handle diverse visual perception tasks. In this paper, we introduce VisionReasoner, a unified framework capable of reasoning and solving multiple visual perception tasks within a shared model. Specifically, by designing a unified reward mechanism and multi-object cognitive learning strategies, VisionReasoner enhances its reasoning capabilities to analyze visual inputs, and addresses diverse perception tasks within a unified model. VisionReasoner generates a structured reasoning process before delivering the desired outputs responding to user queries. Human evaluation reveals the reasoning process of VisionReasoner is faithful and reliable even without annotated reasoning train data. To rigorously assess unified visual perception capabilities, we evaluate VisionReasoner on ten diverse tasks spanning three critical domains: detection, segmentation, and counting. Experimental results show that VisionReasoner achieves superior performance as a unified model, outperforming the baseline Qwen2.5VL by relative margins of 29.1\% on COCO (detection), 22.1\% on ReasonSeg (segmentation), and 13.2\% on CountBench (counting).