A continuous wavelet transform-enhanced transformer model for spectroscopic detection of bacterial wilt in tomato at early infection stages
Xiaoya Ru, Tianzhi He, Huaiheng Cai, Ran Liu, Jiaming Si, Yunming Xu, Yimin Zhou, Zhujun Zhu, Tao Ji · 2026
Read A continuous wavelet transform-enhanced transformer model for spectroscopic detection of bacterial wilt in tomato at early infection stages by Xiaoya Ru, Tianzhi He, Huaiheng Cai, Ran Liu, Jiaming Si, Yunming Xu, Yimin Zhou, Zhujun Zhu, Tao Ji (2026) 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-13582588.txt.
Description
Xiaoya Ru and co-authors present a detection model in two thousand twenty-six for bacterial wilt in tomato at its earliest stages. The soil-borne disease devastates crops, yet early infection shows no visible signs, so ordinary checks miss the window for action. Light spectra from the canopy may betray the disease before eyes can.
Their model fuses raw hyperspectral readings with multi-scale features drawn by a wavelet transform, then reads them with a transformer network. Severity ran from grade zero to four, with grades one and two counted as early. Overall accuracy reached about ninety-seven percent, and early-stage accuracy about ninety-four percent, beating models fed raw spectra alone. Wavelet features lifted every rival model too. An explanation tool highlighted which wavelengths the model relies on.
Catalog record
| CREATOR | Xiaoya Ru, Tianzhi He, Huaiheng Cai, Ran Liu, Jiaming Si, Yunming Xu, Yimin Zhou, Zhujun Zhu, Tao Ji |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 65 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |