Brood Parasitism Identification Using a Deep Learning Model with Mish Activation Function

Scientists are very interested in the phenomenon of bird brood parasitism because it allows them to identify bird species in the nest without having to crack the eggs, which is a destructive process. The host female leaves the nest when she discovers a parasite egg there. Due to this behavior, an important number of future birds disappear and the behavior is extended. Indeed, the cuckoo is one of the species responding to the brood parasitism. Advances in deep learning models in classification and pattern recognition, which are less invasive and can identify individuals based on visual information in the egg image, are discussed in this paper. In this paper, we have suggested a system that automatically extracts the egg’s visual features and classifies them using a CNN model. Additionally, a Mish activation function is introduced to improve extracting depth characteristics to obtain extremely accurate parasitic egg identification. Actually, our model has been evaluated on two data sets: the first is for the GRW egg and Cuckoo egg, and the second is for the RW and Cuckoo egg, and it has an accuracy of 94.7% and 90.9%, respectively.