About TheFly

TheFly was developed by Zenith meditouch in collaboration with:

Nophawan Bunchu, Naresuan University
Paisarn Muneesawang, Naresuan University
Darlin Apasarawirote, Naresuan University
Pharinya, Boonchai, Naresuan University

Support

TheFly was funded by Naresuan University and developed with data primarily from Fly Research NU team, Department of Microbiology and Parasitology, Faculty of Medical Science, Naresuan University.

The Model

The current version of TheFly is the first version and can identify 4 species of the most common fly species found in Thailand including Chrysomya megacephala, Chrysomya (Achoetandrus) rufifacies, Lucilia cuprina, and Musca domestica. This version can identify species of fly larvae by using the images of posterior spiracles. TheFly uses a convolutional neural network, modified from AlexNet, and was trained on over 12,000 images.