PSMamba: Progressive self-supervised vision Mamba for plant disease recognition
- 1Imaging and Computer Vision Group, Data61, CSIRO, Black Mountain, 2601, Canberra, Australia
- 2School of Engineering and Built Environment, Griffith University, Nathan, 4111, Queensland, Australia
- 3School of Information and Communication Technology, Griffith University, Nathan, 4111, Queensland, Australia
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 representative CNN-, Transformer-, SSL-, and Mamba-based baselines, delivering superior accuracy and robustness in both domain-shifted and fine-grained scenarios.
Related
- MambaSeed: A lightweight Mamba-based hybrid vision network for tomato leaf disease recognition
- Neural evidence for progressive decline of visual face recognition ability in Alzheimer’s disease
- Lightweight frequency-aware hybrid vision transformer for fine-grained plant disease recognition
- Comparative longitudinal evaluation of DIMS, progressive addition and single-vision lenses on myopia progression, binocular and visual performance in Spanish children
- Event-driven neuromorphic vision enables energy-efficient visual place recognition
BibTeX
@article{mamun2026psmamba,
title = {{PSMamba: Progressive self-supervised vision Mamba for plant disease recognition}},
author = {Abdullah Al Mamun and Miaohua Zhang and David Ahmedt‐Aristizabal and Zeeshan Hayder and Mohammad Awrangjeb},
journal = {Expert Systems with Applications},
year = {2026},
month = jul,
doi = {10.1016/j.eswa.2026.133775},
url = {https://doi.org/10.1016/j.eswa.2026.133775},
note = {CC BY},
}
RIS
TY - JOUR
TI - PSMamba: Progressive self-supervised vision Mamba for plant disease recognition
AU - Abdullah Al Mamun
AU - Miaohua Zhang
AU - David Ahmedt‐Aristizabal
AU - Zeeshan Hayder
AU - Mohammad Awrangjeb
PY - 2026
DA - 2026/07/30
JO - Expert Systems with Applications
DO - 10.1016/j.eswa.2026.133775
UR - https://doi.org/10.1016/j.eswa.2026.133775
AB - 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 representative CNN-, Transformer-, SSL-, and Mamba-based baselines, delivering superior accuracy and robustness in both domain-shifted and fine-grained scenarios.
SN - 0957-4174
SN - 1873-6793
ER -