Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

InsectMamba: Insect Pest Classification with State Space Model

Qianning Wang, Chenglin Wang, Zhixin Lai, Yucheng Zhou

TL;DR

This work tackles insect pest classification under challenging conditions of camouflage and species diversity. It introduces InsectMamba, a vision model that blends State Space Models with CNNs, Multi-Head Self-Attention, and MLPs inside Mix-SSM blocks, complemented by a selective module that adaptively fuses features from multiple encodings. The key contributions include the first application of SSM-based architectures to insect pest classification, the Mix-SSM Block design for cross-encoding feature extraction, and a selective aggregation mechanism validated across five datasets with comprehensive ablations. The approach yields state-of-the-art performance and demonstrates robustness across varied field conditions, offering practical benefits for pest management and sustainable agriculture.

Abstract

The classification of insect pests is a critical task in agricultural technology, vital for ensuring food security and environmental sustainability. However, the complexity of pest identification, due to factors like high camouflage and species diversity, poses significant obstacles. Existing methods struggle with the fine-grained feature extraction needed to distinguish between closely related pest species. Although recent advancements have utilized modified network structures and combined deep learning approaches to improve accuracy, challenges persist due to the similarity between pests and their surroundings. To address this problem, we introduce InsectMamba, a novel approach that integrates State Space Models (SSMs), Convolutional Neural Networks (CNNs), Multi-Head Self-Attention mechanism (MSA), and Multilayer Perceptrons (MLPs) within Mix-SSM blocks. This integration facilitates the extraction of comprehensive visual features by leveraging the strengths of each encoding strategy. A selective module is also proposed to adaptively aggregate these features, enhancing the model's ability to discern pest characteristics. InsectMamba was evaluated against strong competitors across five insect pest classification datasets. The results demonstrate its superior performance and verify the significance of each model component by an ablation study.

InsectMamba: Insect Pest Classification with State Space Model

TL;DR

This work tackles insect pest classification under challenging conditions of camouflage and species diversity. It introduces InsectMamba, a vision model that blends State Space Models with CNNs, Multi-Head Self-Attention, and MLPs inside Mix-SSM blocks, complemented by a selective module that adaptively fuses features from multiple encodings. The key contributions include the first application of SSM-based architectures to insect pest classification, the Mix-SSM Block design for cross-encoding feature extraction, and a selective aggregation mechanism validated across five datasets with comprehensive ablations. The approach yields state-of-the-art performance and demonstrates robustness across varied field conditions, offering practical benefits for pest management and sustainable agriculture.

Abstract

The classification of insect pests is a critical task in agricultural technology, vital for ensuring food security and environmental sustainability. However, the complexity of pest identification, due to factors like high camouflage and species diversity, poses significant obstacles. Existing methods struggle with the fine-grained feature extraction needed to distinguish between closely related pest species. Although recent advancements have utilized modified network structures and combined deep learning approaches to improve accuracy, challenges persist due to the similarity between pests and their surroundings. To address this problem, we introduce InsectMamba, a novel approach that integrates State Space Models (SSMs), Convolutional Neural Networks (CNNs), Multi-Head Self-Attention mechanism (MSA), and Multilayer Perceptrons (MLPs) within Mix-SSM blocks. This integration facilitates the extraction of comprehensive visual features by leveraging the strengths of each encoding strategy. A selective module is also proposed to adaptively aggregate these features, enhancing the model's ability to discern pest characteristics. InsectMamba was evaluated against strong competitors across five insect pest classification datasets. The results demonstrate its superior performance and verify the significance of each model component by an ablation study.
Paper Structure (23 sections, 16 equations, 5 figures, 7 tables)

This paper contains 23 sections, 16 equations, 5 figures, 7 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. Image Classification
  4. Insect Pest Classification
  5. Preliminaries
  6. Convolutional Neural Networks
  7. Multi-Head Self-Attention
  8. Multi-Layer Perceptron
  9. State Space Models
  10. InsectMamba
  11. Overall Architecture
  12. Mix-SSM Block
  13. Selective Module
  14. Experiment
  15. Dataset and Metrics
  16. ...and 8 more sections

Figures (5)

  • Figure 1: The overall architecture of our InsectMamba Model.
  • Figure 2: Details of our Mix-SSM Block.
  • Figure 3: Comparison of feature aggregation methods for different encoding blocks.
  • Figure 4: Investigation of different kernel sizes in the selective module.
  • Figure 5: Impact of different pooling methods for selective weight generation in the selective module.