Pest-Thinker: 通过强化学习学习像昆虫学家一样思考和推理
Pest-Thinker: Learning to Think and Reason like Entomologists via Reinforcement Learning
AI总结:
本文提出Pest-Thinker框架,通过强化学习提升多模态大语言模型对害虫形态的细粒度理解,构建了两个高精度害虫基准数据集,并通过监督微调和GRPO优化提升模型在农业害虫识别中的表现。
摘要:
害虫引起的作物损失对全球粮食安全和可持续农业发展构成重大威胁。尽管多模态大语言模型(MLLMs)在视觉理解和智能农业方面显示出强大潜力,但其直接应用于害虫识别仍受限于领域特有的挑战,如物种间复杂性、物种内变异性以及专家标注数据稀缺。本文介绍Pest-Thinker,一种知识驱动的强化学习(RL)框架,使MLLMs能够对细粒度害虫形态进行推理。我们首先构建了两个高清晰度害虫基准,QFSD和AgriInsect,包含多样化的物种和专家标注的形态特征。利用这些数据集,我们合成Chain-of-Thought(CoT)推理轨迹,通过监督微调(SFT)促进对害虫特定视觉线索的结构化学习。随后,我们采用具有新特征奖励的组相对策略优化(GRPO),引导模型关注可观察的形态证据,通过LLM-as-a-Judge策略评估。大量实验表明,Pest-Thinker显著提高了领域内和领域外的形态理解能力,标志着向专家级视觉推理迈进的一步。数据集和源代码将在接受后提供。
英文摘要:
Pest-induced crop losses pose a major threat to global food security and sustainable agricultural development. While recent advances in Multimodal Large Language Models (MLLMs) have shown strong potential for visual understanding and smart agriculture, their direct application to pest recognition remains limited due to the domain's unique challenges such as high inter-species complexity, intra-species variability, and the scarcity of expert-annotated data. In this work, we introduce Pest-Thinker, a knowledge-driven reinforcement learning (RL) framework that enables MLLMs to reason over fine-grained pest morphology. We first construct two high-definition pest benchmarks, QFSD and AgriInsect, comprising diverse species and expert-annotated morphological traits. Leveraging these datasets, we synthesize Chain-of-Thought (CoT) reasoning trajectories to facilitate structured learning of pest-specific visual cues through Supervised Fine-Tuning (SFT). Subsequently, we employ Group Relative Policy Optimization (GRPO) with a novel feature reward that guides the model to focus on observable morphological evidence, assessed by an LLM-as-a-Judge strategy. Extensive experiments demonstrate that Pest-Thinker substantially improves both in-domain and out-of-domain morphological understanding, marking a step toward expert-level visual reasoning for intelligent agricultural pest analysis. The datasets and source code are available upon acceptance.
机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。