Expanding the Capacities of a CNN-Based Ant Counting and Density Estimation Application

Understanding collective insect behaviour is a pertinent ecological problem that has to be solved. The fact that some of these flocks are the largest cooperative units in nature has a tremendous impact on our knowledge of nature. Reliable data sampling contributes to the challenge’s complexity in some ways. Based on convolutional neural networks, this paper proposes a novel technique for comprehending the numbers and distribution of ants in colonies. Due to the special nature of this tool, we developed an application to generate the marked dataset, produced the initial dataset, and evaluated the solution using several backbones. Our findings imply that the suggested strategy is workable to address the suggested problem. The average coefficient of determination \(R^2\) with the ground truth counting was 0.9783 using the MobileNet as the backbone and 0.9792 using the EfficientNet-V2B0 as the backbone. The global average for the semi-quantitive classification of each image region was 86% for the MobileNet and 88% for the EfficientNet V2-B0. There was no statistically significant difference between both cases’ average and median errors. The coefficient of determination was close to the statistical significance threshold ( \(p = 0.065\) ). The application using the MobileNet as its backbone performed the task faster than the version using the EfficientNet V2-B0, with statistical significance ( \(p < 0.05\) ).