Potato defect detection is a critical step in post-harvest sorting. Traditional supervised methods rely on large manually labeled datasets, which is time-consuming and labour-intensive. This study proposes a novel semi-supervised learning (SSL) approach for potato defect detection. We first compared multiple neural networks and selected ConvNeXt v2 as the backbone. We then evaluated various SSL architectures and selected FixMatch. The proposed ConvNeXt v2 + FixMatch framework achieved superior performance across all experiments. With only 50, 100, 200, and 400 labeled images, the test accuracies reached 0.880, 0.899, 0.937, and 0.943, respectively, outperforming fully supervised learning (FSL) baselines by 0.117, 0.118, 0.120, and 0.050. Furthermore, the model trained with limited labeled samples can serve as a seed model to automatically annotate unlabeled samples, enabling large-scale dataset labeling. These results demonstrate that SSL methods effectively leverage unlabeled images to enhance neural network performance for potato defect detection.
Chen, Fengnong, Xiang, Yuheng, Li, You et al. Different Learning Model Fusion Method under Small Labeled Samples for Potatoes External Defect Detection. // Potato Research. – 2026. – DOI: 10.1007/s11540-026-10147-3. – URL: http://link.springer.com/openurl/fulltext?id=doi:10.1007/s11540-026-10147-3
Chen, Fengnong, Xiang, Yuheng, Li, You, Chen, Zhaoqing, Wei, Qiquan, Sun, & Hongwei (2026). Different Learning Model Fusion Method under Small Labeled Samples for Potatoes External Defect Detection. Potato Research. https://doi.org/10.1007/s11540-026-10147-3
Chen, et al. “Different Learning Model Fusion Method under Small Labeled Samples for Potatoes External Defect Detection.” Potato Research, 2026. doi:10.1007/s11540-026-10147-3.
@article{Chen2026,
title = {Different Learning Model Fusion Method under Small Labeled Samples for Potatoes External Defect Detection},
author = {Chen and Fengnong and Xiang and Yuheng and Li and You and Chen and Zhaoqing and Wei and Qiquan and Sun and Hongwei},
journal = {Potato Research},
year = {2026},
doi = {10.1007/s11540-026-10147-3},
url = {http://link.springer.com/openurl/fulltext?id=doi:10.1007/s11540-026-10147-3}
}
Soil-related properties associated with potato yield and quality are implicitly linked to soil water storage (SWS) determined by soil texture (ST). This soil property also affects the potato canopy, which could be evaluated indirectly...
Soft rot and blackleg diseases caused by Soft Rot Pectobacteriaceae (SRP) continue to significantly hinder global potato production. However, their epidemiology is often attributed to a single pathogen, which oversimplifies their...
A pesar de la larga historia del cultivo de papa en Brasil, su dinámica territorial sigue siendo poco comprendida. Analizamos datos de producción municipal y macrorregional de 1994 a 2024 utilizando análisis de ruptura estructural,...
The experiments were conducted at the tissue culture laboratory of the National Commission for Biotechnology (NCBT) and the Faculty of Science, Damascus University, to evaluate the response of nine introduced potato cultivars (...