Synopsis

Explores the application of Few-Shot Learning (FSL) and Zero-Shot Learning (ZSL) in computer vision, particularly for agricultural and medical image analysis, where obtaining large, accurately annotated datasets can be expensive, time-consuming, and challenging.

Relevance of the Topic

• Reduces dependence on large, expert-annotated datasets.
• Enables recognition of rare and previously unseen diseases.
• Addresses class imbalance and limited training data.
• Improves model generalization across datasets, imaging conditions, and environments.

Future Research/Scope

• Few-Shot and Zero-Shot Classification: Recognition of rare or previously unseen agricultural and medical diseases.
• Few-Shot and Zero-Shot Segmentation: Localization of plant infections, tumors, and lesions.
• Anomaly Detection: Identification of abnormalities using predominantly normal images.

Skills Learned

• Computer Vision: Image classification, segmentation, and anomaly detection.
  Deep Learning: Transfer learning, self-supervised learning, and foundation models 
  Few-Shot and Zero-Shot Learning: Metric learning, prompt learning, and representation learning.

Relevant courses to the topic

• CSE 425: Neural Networks
• CSE 427: Machine Learning
• CSE 428: Image Processing

Reading List

• Ge, Z., Fan, X., Zhang, J., & Jin, S. (2025). SegPPD-FS: Segmenting plant pests and diseases in the wild using few-shot learning. Plant Phenomics, 7(4), 100121. https://doi.org/10.1016/j.plaphe.2025.100121 
•  Shiri, M., Beyan, C., & Murino, V. (2025). MadCLIP: Few-shot medical anomaly detection with CLIP. arXiv. https://doi.org/10.48550/arXiv.2506.23810 
• Khankalantary, S., & Eskandari, M. R. (2026). Dual-stage parameter-efficient tuning for in-field plant disease few-shot classification. Computers and Electronics in Agriculture, 248, 111691. https://doi.org/10.1016/j.compag.2026.111691 
• Jiang X, Wang J, Xie K, Cui C, Du A, Shi X, Yang W, Zhai R. PlantCaFo: An efficient few-shot plant disease recognition method based on foundation models. Plant Phenomics. 2025 Feb 28;7(1):100024. doi: 10.1016/j.plaphe.2025.100024. PMID: 41415957; PMCID: PMC12709961. https://pmc.ncbi.nlm.nih.gov/articles/PMC12709961/