Artificial Intelligence Analysis of Dietary Adaptation Among Urban Birds

Abstract

Introduction: Urbanization has drastically impacted natural ecosystems in terms of availability of food, habitat alteration, and resource availability, thus compelling birds to alter their food habits. Birds in cities rely on multiple food items, including natural food as well as food obtained from humans. It is important to understand dietary adaptations of birds to be able to evaluate survival strategies of various species along with changes in urban biodiversity. Dietary analysis through conventional techniques such as direct observations in the field is quite cumbersome and inefficient in understanding behavioural changes at a broader scale.

Objective: The objective of this study is to present a machine learning-based approach for dietary adaptations of urban birds through computer vision and deep learning.

Methodology: The methodological framework includes pre-processing of images, a deep-learning driven bird detection system, and a classification of their behaviors to determine how birds adapt to their diets in different urban environments. The acquired data are classified according to the bird species, their feeding behavior, and food sources. The trained model determines how there is a variance in diet adaptation in different urban bird populations.

Results: In experimental testing of the methodological framework, it was found out that the artificial intelligence model successfully identifies the bird species and classifies their behavior with increased precision. There is variance in food selection among different urban bird populations depending on the resources available.

Conclusion: The presented methodological framework based on artificial intelligence can provide a solution to automatically monitor how birds adapt to their diets in different urban areas.