Please login first
Systematics of Tephritid Fruit Flies: A Machine Learning Based Pest Identification System
1 , * 2, 3 , 4 , * 5
1  Research School of Computing, The Australian National University (ANU), Canberra, ACT 2601, Australia
2  CSIRO Agriculture and Food, Black Mountain, Canberra, ACT 2601, Australia
3  Research School of Biology and Research School of Computing, ANU, Canberra, ACT 2601, Australia
4  Department of Electrical and Computer Engineering, North South University, Dhaka, Bangladesh
5  CSIRO Land and Water, Black Mountain, Canberra, ACT 2601, Australia
Academic Editor: Andjeljko Petrović


Tephritid fruit flies (Diptera: Tephritidae) are the major economically important agricultural pests around the world. Numerous control measures are undergoing to reduce their abundance. An efficient pest identification system is a prerequisite for such tasks. Typically, the classification/identification of different insect species is done based on either external body features or DNA barcoding. However, those approaches are time-consuming by nature, requiring expert knowledge in relevant fields. Several machine learning (ML) models have been successfully deployed in the field of systematics, but there is a lack of ML models for fruit fly species. This study aims to curate and validate a comprehensive tephritid image database and build ML models to automatically identify Tephritids from non-Tephritid dipteran flies and classify four major genera of notorious Tephritid flies, namely, Anastrepha, Ceratitis, Rhagoletis, and Bactrocera. The images of our experiment were collected from the iNaturalist database. The dataset was cleaned by removing uninformative images using a deep learning model (Inception-V3) and unsupervised k-mean clustering. Several state-of-the-art ML models were tested on the dataset, resulted in the highest accuracy of 95.44% with the EfficientNet-B0 model to identify tephritid flies from non-tephritids. Moreover, the EfficientNet-B2 model achieved 88.68% accuracy for classifying representatives of the major tephritid genus and showed the potential to enhance the identification accuracy. Overall, this work of the systematics of harmful fruit flies can be transformed into a practical and effective detection tool and can be implemented easily with existing agricultural pest control systems.

Keywords: Tephritid fruit fly; Machine Learning; Artificial Intelligence; Insect systematics