Improved CSW-YOLO Model for Bitter Melon Phenotype Detection
Haobin Xu, Xianhua Zhang, Weilin Shen, Zhiqiang Lin, Shuang Liu, Qi Jia, Honglong Li, Jingyuan Zheng, Fenglin Zhong, Georgios Koubouris · 2024
Read Improved CSW-YOLO Model for Bitter Melon Phenotype Detection by Haobin Xu, Xianhua Zhang, Weilin Shen, Zhiqiang Lin, Shuang Liu, Qi Jia, Honglong Li, Jingyuan Zheng, Fenglin Zhong, Georgios Koubouris (2024) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/pmc-11644074.txt.
Description
Haobin Xu and co-authors build a vision model that identifies bitter melon types from photographs. The gourd's many shapes affect its market value and breeding value, yet sorting germplasm by eye is slow and error-prone. Their model, CSW-YOLO, upgrades a popular detection network for this farm task.
Three changes drive the upgrade: a stronger feature extractor, an attention trick that highlights key neurons at no extra cost, and a better training target for tight boxes. On their melon image set, the model reached about ninety-five percent precision and a top detection score near ninety-seven percent. It beat the base network on every measure and outperformed other mainstream models in side-by-side tests. Heat maps confirmed the model looks at the fruit itself rather than the background. The authors offer the tool for breeding programs and, more broadly, for automated crop inspection.
Catalog record
| CREATOR | Haobin Xu, Xianhua Zhang, Weilin Shen, Zhiqiang Lin, Shuang Liu, Qi Jia, Honglong Li, Jingyuan Zheng, Fenglin Zhong, Georgios Koubouris |
|---|---|
| YEAR | 2024 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 35 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |