Perceive-to-Reason: Decoupling Perception and Reasoning for Fine-Grained Visual Reasoning
cs.CV
Submitted: 2026-07-01
Updated: 2026-11-16
Comments: Code: https://github.com/ZJU-REAL/Perceive-to-Reason
License: http://creativecommons.org/licenses/by/4.0/
The gist: Fine-grained visual reasoning remains challenging for vision-language models, especially when small but critical visual cues are buried in high-resolution images.
Terminology
Abstract
Fine-grained visual reasoning remains challenging for vision-language models, especially when small but critical visual cues are buried in high-resolution images. Existing approaches rely on repeated cropping or test-time visual search to introduce local evidence, but they typically do not explicitly distinguish perception from reasoning. In this paper, we propose Perceive-to-Reason (P2R), a unified framework that formulates fine-grained visual reasoning as a two-stage process: the model first localizes question-relevant evidence as a Perceiver, and then answers the question as a Reasoner based on the annotated image and cropped regions. To better align training with this decoupled formulation, we further introduce Perception-Reasoning Alternating GRPO (PRA-GRPO), a role-aware reinforcement learning strategy that alternates between perception-focused and reasoning-focused updates using only final-answer supervision. Built on top of Qwen3-VL-Instruct-2B/4B/8B, P2R consistently improves performance across model scales. In particular, P2R-4B achieves 93.2% on V-Star, 81.9% on HR-Bench-4K, and 80.5% on HR-Bench-8K, substantially outperforming its corresponding backbone. Further experiments show that the benefits of P2R extend beyond high-resolution benchmarks to broader multimodal reasoning tasks. These results suggest that explicitly decoupling perception from reasoning provides an effective framework for fine-grained visual reasoning.
Sources
- Qwen3-VL Technical Report
- Perception Before Reasoning: Two-Stage Reinforcement Learning for Visual Reasoning in Vision-Language Models
- GRIT: Teaching MLLMs to Think with Images
- SRFT: A Single-Stage Method with Supervised and Reinforcement Fine-Tuning for Reasoning
- DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning
- DeepEyesV2: Toward Agentic Multimodal Model
- Vision-R1: Incentivizing Reasoning Capability in Multimodal Large Language Models
- GPT-4o System Card
- Mini-o3: Scaling Up Reasoning Patterns and Interaction Turns for Visual Search
- ViewSpatial-Bench: Evaluating Multi-perspective Spatial Localization in Vision-Language Models
- Reliable Thinking with Images
- SpatialLadder: Progressive Training for Spatial Reasoning in Vision-Language Models
- VLM-FO1: Bridging the Gap Between High-Level Reasoning and Fine-Grained Perception in VLMs
- GDPO: Group reward-Decoupled Normalization Policy Optimization for Multi-reward RL Optimization
- HiDe: Rethinking The Zoom-IN method in High Resolution MLLMs via Hierarchical Decoupling
- Seg-Zero: Reasoning-Chain Guided Segmentation via Cognitive Reinforcement
- Understanding R1-Zero-Like Training: A Critical Perspective
- Visual-RFT: Visual Reinforcement Fine-Tuning
- Chain-of-Spot: Interactive Reasoning Improves Large Vision-Language Models
- Proximal Policy Optimization Algorithms
Related papers
- TEA: Text Encoder Alignment for Robust Concept Erasure in Text-to-Image Models
- Do Egocentric Video-Language Models Capture Both Hand- and Object-Centric Cues?
- Aura: Consistent Multi-Subject Video Generation via VLM-Grounded Semantic Alignment
- PVRF: All-in-one Adverse Weather Removal via Prior-modulated and Velocity-constrained Rectified Flow
- DLWM: Diverse Latent World Models for Efficient Multimodal Reasoning
- Out of Sight, Still in Mind: Token Compression for Omni-LLMs