Related Experiment Video
Updated: May 1, 2026

Comparison of Methods for Isolating Entomopathogenic Fungi from Soil Samples
Published on: January 6, 2022
Innovative infrastructure to access Brazilian fungal diversity using deep learning
Thiago Chaves1, Joicymara Santos Xavier2, Alfeu Gonçalves Dos Santos1
1Brazilian National Institute for Digital Convergence-INCoD, Universidade Federal de Santa Catarina, Florianópolis, Santa Catarina, Brazil.
This study introduces a new database of Brazilian macrofungi to train convolutional neural networks (CNNs) for automated species identification, enabling a mobile app for public use and fungal conservation efforts.
Area of Science:
- Mycology
- Computer Science
- Biodiversity Informatics
Background:
- Accurate identification of macrofungal species is crucial for ecological studies and conservation.
- Existing methods for macrofungal identification can be time-consuming and require expert knowledge.
- There is a need for accessible tools to aid in the identification and documentation of Brazilian fungi.
Purpose of the Study:
- To develop a comprehensive, expert-curated database of Brazilian macrofungi with extensive photographic data.
- To train and validate convolutional neural networks (CNNs) for automated macrofungal species identification.
- To create a user-friendly mobile application for image-based macrofungal identification, promoting public engagement and data collection.
Main Methods:
- Compilation of a meticulously structured database of macrofungi from Brazil, including over 13,894 photographs of 505 species.
- Training and validation of convolutional neural networks (CNNs) using the curated image database for autonomous species recognition.
- Development of a mobile application with an advanced user interface for image acquisition and AI-driven identification suggestions.
Main Results:
- Successful creation of a large-scale, expert-verified database of Brazilian macrofungi.
- Demonstrated capability of trained CNNs to autonomously identify macrofungal species from images.
- Development of a functional mobile application prototype for public use.
Conclusions:
- The developed database and CNN models provide a powerful tool for macrofungal identification in Brazil.
- The mobile application democratizes access to knowledge about Brazilian fungi, fostering public engagement and citizen science.
- This technology supports enhanced biodiversity monitoring and conservation efforts for Brazilian macrofungi.
More Related Videos
11:48Microscopy Techniques for Interpreting Fungal Colonization in Mycoheterotrophic Plants Tissues and Symbiotic Germination of Seeds
Published on: May 17, 2022
07:00Microbiota of Attine Ants' Gardens: Visualizing a Microbial Landscape by Scanning Electron Microscopy
Published on: October 4, 2024
Related Concept Videos
Modern Molecular Taxonomy
Applications of Molecular Taxonomy
Other Unique Bacteria
Methods to Assess Microbial Communities
Development of Human Microbiota
Automated Microbial Diagnostics