EUVIMEDObjevte EUVIMED
Crossref

Causality‐inspired crop pest recognition based on Decoupled Feature Learning

Tao Hu iD, Jianming Du, Keyu Yan, Wei Dong, Jie Zhang, Jun Wang, Chengjun Xie

Pest Management Science · 2024-11 · 80 (11) · 5832-5842

DOI: 10.1002/ps.8314 ↗

At a glance

Journal article in Pest Management Science, 2024-11. Written by Tao Hu and 6 co-authors. Institution: Science Island Branch of Graduate School, University of Science and Technology of China, Hefei, China. Major MEDLINE topics: Crops, Agricultural, Deep Learning. Cited 10 times so far (Crossref); 32 references.

AbstraktSubjectsBibliografické údajeCiteRelated studies

Abstract (original language)

BACKGROUND Ensuring the efficient recognition and management of crop pests is crucial for maintaining the balance in global agricultural ecosystems and ecological harmony. Deep learning‐based methods have shown promise in crop pest recognition. However, prevailing methods often fail to address a critical issue: biased pest training dataset distribution stemming from the tendency to collect images primarily in certain environmental contexts, such as paddy fields. This oversight hampers recognition accuracy when encountering pest images dissimilar to training samples, highlighting the need for a novel approach to overcome this limitation. RESULTS We introduce the Decoupled Feature Learning (DFL) framework, leveraging causal inference techniques to handle training dataset bias. DFL manipulates the training data based on classification confidence to construct different training domains and employs center triplet loss for learning class‐core features. The proposed DFL framework significantly boosts existing baseline models, attaining unprecedented recognition accuracies of 95.33%, 92.59%, and 74.86% on the Li, DFSPD, and IP102 datasets, respectively. CONCLUSION Extensive testing on three pest datasets using standard baseline models demonstrates the superiority of DFL in pest recognition. The visualization results show that DFL encourages the baseline models to capture the class‐core features. The proposed DFL marks a pivotal step in mitigating the issue of data distribution bias, enhancing the reliability of deep learning in agriculture. © 2024 Society of Chemical Industry.

Subjects

MeSH terms (MEDLINE)

Crops, AgriculturalDeep LearningAnimalsPest Control

Author keywords

Decoupled Feature Learningcausal inferencedeep learningpest recognition

Authors and institutions

  • Science Island Branch of Graduate School, University of Science and Technology of China, Hefei, China.
  • Institute of Intelligent Machines, Hefei Institutes of Physical Science, Chinese Academy of Sciences, Hefei, China.
  • Agricultural Economy and Information Research Institute, Anhui Academy of Agricultural Sciences, Hefei, China.
  • School of Internet and Telecommunication, Anhui Technical College of Mechanical and Electrical Engineering, Wuhu, China.

ORCID profiles: Tao Hu

Cite

APA

Hu, T., Du, J., Yan, K., Dong, W., Zhang, J., Wang, J., & Xie, C. (2024). Causality‐inspired crop pest recognition based on Decoupled Feature Learning. Pest Management Science. https://doi.org/10.1002/ps.8314

Vancouver

Hu T, Du J, Yan K, Dong W, Zhang J, Wang J, et al. Causality‐inspired crop pest recognition based on Decoupled Feature Learning. Pest Management Science. 2024. doi:10.1002/ps.8314

BibTeX
@article{Hu2024,
  author = {Hu, Tao and Du, Jianming and Yan, Keyu and Dong, Wei and Zhang, Jie and Wang, Jun and Xie, Chengjun},
  title = {Causality‐inspired crop pest recognition based on Decoupled Feature Learning},
  journal = {Pest Management Science},
  year = {2024},
  doi = {10.1002/ps.8314}
}
Start for free →