Machine Learning Approaches for Automated Classification and Identification of Bird Vocalizations

Authors

DOI:

https://doi.org/10.17010/ijcs/2026/v11/i5/176125

Keywords:

Bioacoustics, bird sound classification, class imbalance, macro-F1, MobileNetV, Random Forest, Support Vector Machine, Transfer learning, YAMNet


Publishing Chronology: Paper Submission Date : August 12, 2026 ; Paper sent back for Revision : August 23, 2026 ; Paper Acceptance Date : August 26, 2026 ; Paper Published Online : October 5, 2026.

Abstract

Manual classification of birds from their voice is a complex, time-consuming, and challenging task. This study compares two transfer learning approaches for bird classification. In the first approach, audio was converted into Mel spectrograms, which were then used to fine-tune MobileNetV2. In the second approach, YAMNet was used as a feature extractor, and embeddings from YAMNet were used to train a support vector machine, multilayer perceptron, and a random forest. Random forest achieved the highest accuracy and 0.654 as its macro F1 score. Kaggle dataset “Sound of 114 Species of Birds till 2022” was used in both approaches. This study concludes that audio-based pretrained models outperform image-based pretrained models on the same dataset. This study performs a comparative analysis of an audio-based model and an image-based model on the same dataset.

Downloads

Download data is not yet available.

Published

2026-10-30

How to Cite

Khare, S., Vandana, & Patel, A. (2026). Machine Learning Approaches for Automated Classification and Identification of Bird Vocalizations. Indian Journal of Computer Science, 11(5), 8–16. https://doi.org/10.17010/ijcs/2026/v11/i5/176125

References

[1] X. Shen, J. Xu, Q. Zhang, P. Luo, Z. Liu, J. Chen, G. Zhou, J. Yi, Z. Xiao, and A. Chen, “Automated bird sound recognition in ecology: A methodological review and application framework,” Avian Res., vol. 17, no. 3, Art. no. 100368, Sep. 2026.

[2] R. Shashidhar, N. G. Nayaka, S. Chethan, and J. Madhura, “Deep learning approaches for automated bird sound classification using MFCC features,” Int. J. Inf. Technol,. pp. 1–17, Dec. 2025, doi: 10.1007/s41870-025-02981-4.

[3] R. Kamarajugadda, R. Battula, C. R. Borra, H. Durga, V. Bypilla, S. S. Reddy, F. F. Khan, and S. Bhavanam, “Optimizing avian species recognition with MFCC features and deep learning models,” Int. J. Inf. Technol., vol. 16, pp. 4621–4626, Aug. 2024, doi: 10.1007/s41870-024-02108-1.

[4] B. Ghani, T. Denton, S. Kahl, and H. Klinck, “Global birdsong embeddings enable superior transfer learning for bioacoustic classification,” Sci. Rep., vol. 13, no. 1, 22876, 2023, doi: 10.1038/s41598-023-49989-z.

[5] M. V. Conde and U.-J. Choi, “Few-shot long-tailed bird audio recognition,” 2022, arXiv:2206.11260. [6] S. Kahl, A. Navine, T. Denton, H. Klinck, P. Hart, H. Glotin, H. Goëau, W.-P. Vellinga, R. Planqué, and A. Joly, “Overview of birdclef 2022: Endangered bird species recognition in soundscape recordings,” in CEUR-WS, 2022.

[7] S. Kahl, C. M. Wood, M. Eibl, and H. Klinck, “BirdNET: A deep learning solution for avian diversity monitoring,” Ecological Inform., 61, 101236, Mar. 2021, doi: 10.1016/j.ecoinf.2021.101236.

[8] B. Ghani, V. J., Kalkman, B. Planqué, W. P. Vellinga, L. Gill, and D. Stowell, “Impact of transfer learning methods and dataset characteristics on generalization in birdsong classification,” Scientific Reports, vol. 15, no. 1, Art. no. 16273, May 2025, doi: 10.1038/s41598-025-00996-2.

[9] H. Dogan, “Improved detection of bird vocalisations using BirdNET embeddings and machine learning,” 2025, doi: 10.31224/4466.

[10] H. A. Endalamaw, H., C.-K. Yang, and C.-H. Hsu, “A real-time bird sound recognition app via deep learning techniques,” Multimedia Tools Appl., vol. 85, Art. no. 62, Jan. 2026.

[11] M. K. Baowaly, B. C. Sarkar, M. A. A. Walid, M. M. Ahamad, B. C. Singh, E. S. Alvarado, I. Ashraf, and M. A. Samad, “Deep transfer learning-based bird species classification using mel spectrogram images,” PLoS One, vol. 19, no. 8, Aug. 2024, doi: 10.1371/journal.pone.0305708. [12] F. Zhang, L. Zhang, H. Chen, and J. Xie, “Bird species identification using spectrogram based on multi-channel fusion of DCNNs,” Entropy, vol. 23, no. 11, Art. no. 1507, 2021, doi: 10.3390/e23111507.

[13] A. Márquez-Rodríguez, M. A., Mohedano-Munoz, M. J. Marín-Jiménez, E. Santamaría-García, G. Bastianelli, P. Jordano, and I. Mendoza, “A bird song detector for improving bird identification through deep learning: A case study from Doñana,” Ecological Inform., 90, Art. no. 103254, Dec. 2025.

[14] T. Dogandzic, M. Zivkovic, L. Jovanovic, J. Perisic, M. Milovanovic, M. Antonijevic, M. Hammad, and N. Bacanin, “Classifying birds of South America via audio analysis using convolutional networks and boosting models optimized by metaheuristics,” Iran J. Comput. Sci., vol. 9, no. 1, Jan. 2026, doi: 10.1007/s42044-025-00363-4.

[15] J. Xie, S. Xie, Y. Liu, X. Jing, M. Zhu, L. Xie, J. Zhang, K. Qian, and B. W.Schuller, “Decoding nature's melody: Significance and challenges of machine learning in assessing bird diversity via soundscape analysis,” Artif. Intell. Rev., vol. 59, no. 10, 2025, doi: 10.1007/s10462-025-11414-4.

[16] Bird-Sound-Classifier. [Online]. Available: https://github.com/sharmax-vandana/Bird-Sound-Classif