Explainable deep learning techniques for microscopic fungi classification using a learnable threshold-based ReLULeaky activation function and transfer learning
Mohsin Hossain, Mohammad Khairul Islam, Machbah Uddin, Farah Jahan, Md Shafiul Alam Chowdhury, Md Ahadullah · 2026
Read Explainable deep learning techniques for microscopic fungi classification using a learnable threshold-based ReLULeaky activation function and transfer learning by Mohsin Hossain, Mohammad Khairul Islam, Machbah Uddin, Farah Jahan, Md Shafiul Alam Chowdhury, Md Ahadullah (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/PMC13382049.txt.
Description
Mohsin Hossain and co-authors present a 2026 computer study on identifying microscopic fungi from images, with explanations for its decisions. Fungal infections are a growing risk, yet diagnosis is slow and specialists are scarce. Early, accurate identification matters for effective treatment.
The team fine-tuned a well-known image network, adding a new activation method whose threshold learns during training. The model classified fungi with accuracy above ninety-five percent in their tests, beating comparison methods. Explanation tools showed which image features drove each call. Such a system could support labs that lack expert staff, provided its performance holds up on independent image sets from other sources.
Catalog record
| CREATOR | Mohsin Hossain, Mohammad Khairul Islam, Machbah Uddin, Farah Jahan, Md Shafiul Alam Chowdhury, Md Ahadullah |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 68627 |
| LICENSE | https://creativecommons.org/licenses/by/4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |