Automatic Fungi Recognition: Deep Learning Meets Mycology

• Published in: Sensors (Basel, Switzerland) Vol. 22; no. 2; p. 633
• Main Authors: Picek, Lukáš, Šulc, Milan, Matas, Jiří, Heilmann-Clausen, Jacob, Jeppesen, Thomas S, Lind, Emil
• Format: Journal Article
• Language: English
• Published: Switzerland MDPI AG 01-01-2022; MDPI
• Subjects: Applications programs; Artificial intelligence; Biodiversity; Classification; Collaboration; computer vision; Datasets; Deep Learning; Endangered & extinct species; Fungi; Humans; Image retrieval; Machine learning; Metadata; Microscopy; Mobile Applications; Mycology; Neural networks; Neural Networks, Computer; Recognition; Science; species; Web applications; Wireless telephone software; Denmark; Germany; fungi; species; computer vision; recognition; species recognition; classification; fine-grained; machine learning; artificial intelligence
• ISSN: 1424-8220; 1424-8220
• DOI: 10.3390/s22020633

The article presents an AI-based fungi species recognition system for a citizen-science community. The system's real-time identification too - FungiVision - with a mobile application front-end, led to increased public interest in fungi, quadrupling the number of citizens collecting data. FungiVision, deployed with a human-in-the-loop, reaches nearly 93% accuracy. Using the collected data, we developed a novel fine-grained classification dataset - Danish Fungi 2020 (DF20) - with several unique characteristics: species-level labels, a small number of errors, and rich observation metadata. The dataset enables the testing of the ability to improve classification using metadata, e.g., time, location, habitat and substrate, facilitates classifier calibration testing and finally allows the study of the impact of the device settings on the classification performance. The continual flow of labelled data supports improvements of the online recognition system. Finally, we present a novel method for the fungi recognition service, based on a Vision Transformer architecture. Trained on DF20 and exploiting available metadata, it achieves a recognition error that is 46.75% lower than the current system. By providing a stream of labeled data in one direction, and an accuracy increase in the other, the collaboration creates a virtuous cycle helping both communities.