AI Imaging is For The Birds

Image: MetaAI, (2025).

This peer-reviewed conference paper discusses the studies on bird species and the valuable information their patterns and cycles provide us with about biological systems. Identification by photography is time-consuming. The paper looks at recent advances in AI and imaging tools, and discusses the goal of developing computer vision tools, naming transfer learning and deep quantum learning as examples.

They state that the model is effective and will form the basis for what they are calling "the most advanced (SOTA) classification of birds, with 88% accuracy. One such use of this is Bird@Edge, which is described as an AI-based instant bird deterrent to protect plants from harmful birds, “for ecological protection and permaculture promotion, plant protection, biodiversity management."

Citation:

Chaturvedi, K., Skantha, M. N., Navanithi, v k, Pranesh, r, Harissh, G., & Tangavel, S. K. (2024). Deep Learning-Based Identification of Birds in Agricultural Fields. 2024 International Conference on Computing and Intelligent Reality Technologies (ICCIRT), Computing and Intelligent Reality Technologies (ICCIRT), 2024 International Conference On, 1–6. https://doi.org/10.1109/ICCIRT59484.2024.10921862