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

CREATORMohsin Hossain, Mohammad Khairul Islam, Machbah Uddin, Farah Jahan, Md Shafiul Alam Chowdhury, Md Ahadullah
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE68627
LICENSEhttps://creativecommons.org/licenses/by/4.0
RIGHTSCC BY
SOURCEEurope PMC

Source: https://europepmc.org/article/MED/PMC13382049

Browse the full library · About the Tower of Babel