PSMamba: Progressive Self-supervised Vision Mamba for Plant Disease Recognition

2025-12-16

AI Summary

Existing self-supervised methods struggle to effectively model multi-scale and hierarchical lesion patterns in plant disease images. To address this, we propose PSMamba, a progressive self-supervised framework. Its core innovation is a novel “shared global teacher with dual specialized students” architecture: one student models mid-scale features (e.g., lesion distribution and vein structure), while the other captures local-scale features (e.g., textural anomalies and early-stage lesions). PSMamba integrates Vision Mamba’s long-range sequence modeling capability, hierarchical knowledge distillation between the two students, multi-view contrastive learning, and a cross-scale consistency loss. Evaluated on three major benchmark datasets, PSMamba significantly outperforms state-of-the-art self-supervised methods—achieving superior accuracy and robustness under domain shift and in fine-grained disease classification tasks.

Technology Category

• Computer Vision: Multi-modal Vision
• Machine Learning: Multi-instance/Multi-view Learning
• Search and Optimization: Learning to Search

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Web Mining and Content Analysis: Large pretrained models with web data
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

Self-supervised Learning (SSL) has become a powerful paradigm for representation learning without manual annotations. However, most existing frameworks focus on global alignment and struggle to capture the hierarchical, multi-scale lesion patterns characteristic of plant disease imagery. To address this gap, we propose PSMamba, a progressive self-supervised framework that integrates the efficient sequence modelling of Vision Mamba (VM) with a dual-student hierarchical distillation strategy. Unlike conventional single teacher-student designs, PSMamba employs a shared global teacher and two specialised students: one processes mid-scale views to capture lesion distributions and vein structures, while the other focuses on local views to capture fine-grained cues such as texture irregularities and early-stage lesions. This multi-granular supervision facilitates the joint learning of contextual and detailed representations, with consistency losses ensuring coherent cross-scale alignment. Experiments on three benchmark datasets show that PSMamba consistently outperforms state-of-the-art SSL methods, delivering superior accuracy and robustness in both domain-shifted and fine-grained scenarios.

Problem

• Proposes PSMamba for hierarchical plant disease recognition
• Addresses multi-scale lesion pattern capture in SSL
• Enhances accuracy in domain-shifted and fine-grained scenarios

Innovation

• Integrates Vision Mamba for efficient sequence modeling
• Uses dual-student hierarchical distillation for multi-scale learning
• Employs consistency losses for cross-scale alignment