Comparative Evaluation of YOLO-family Detectors for Pig Detection in Precision Livestock Systems

Marcos Vinicius Mendes Faria; Thiago Meireles Paixão; Francisco de Assis Boldt

doi:10.22456/2175-2745.150940

Authors

Marcos Vinicius Mendes Faria Instituto Federal de Educação, Ciência e Tecnologia do Espírito Santo (IFES) https://orcid.org/0009-0005-7970-2228
Thiago Meireles Paixão Instituto Federal de Educação, Ciência e Tecnologia do Espírito Santo (IFES) https://orcid.org/0000-0003-1554-6834
Francisco de Assis Boldt Instituto Federal de Educação, Ciência e Tecnologia do Espírito Santo (IFES) https://orcid.org/0000-0001-6919-5377

DOI:

https://doi.org/10.22456/2175-2745.150940

Keywords:

precision livestock farming, object detection, deep learning, YOLOv8, YOLOv12, pig monitoring, computer vision

Abstract

Precision Livestock Farming (PLF) delivers technological solutions for modern pork production, where real-time monitoring is vital for animal health and welfare. Object detection models are key in PLF, yet comprehensive baselines for recent architectures on specialized datasets remain limited. This study benchmarks two state-of-the-art models, YOLOv8 and YOLOv12, across five sizes (nano to extra-large) for pig detection using the public PigLife dataset. Ten models were fine-tuned and evaluated with standard COCO metrics. Results show that while accuracy improves with model size, gains diminish with larger models. The YOLOv8m, with 25.9 million parameters, achieved a near-peak mean Average Precision (mAP) of 91.8%. This work establishes a new benchmark on PigLife, surpassing previous results, and highlights the trade-off between YOLOv8’s higher accuracy and YOLOv12’s parameter efficiency.

Downloads

Download data is not yet available.

References

[1] FAO. FAO Knowledge Repository. 2025. Accessed: 2025-08-15. Disponível em: https://openknowledge.fao.org/items/b553c447-f6b2-4ff8-825e-024767b27760.

[2] Crop Biotech Update. FAO: The World Needs 70% More Food by 2050. 2024. Accessed: 2025-08-15. Disponível em: https://www.isaaa.org/kc/cropbiotechupdate/article/default.asp?ID=4778.

[3] BERCKMANS, D. General introduction to precision livestock farming. Animal Frontiers, v. 7, n. 1, p. 6–11, Jan 2017.

[4] TAGOE, S. N. A.; ELMIR, M.; AMANQUAH, N. Monitoring of animal movement using computer vision. In: 2024 IEEE 9th International Conference on Adaptive Science and Technology (ICAST). Accra, Ghana: [s.n.], 2024. p. 1–6.

[5] LI, J. et al. Barriers to computer vision applications in pig production facilities. Computers and Electronics in Agriculture, v. 200, p. 107227, 2022.

[6] VARGHESE, R.; S, M. Yolov8: A novel object detection algorithm with enhanced performance and robustness. In: [S.l.: s.n.], 2024. p. 1–6.

[7] TIAN, Y.; YE, Q.; DOERMANN, D. YOLOv12: Attention-Centric Real-Time Object Detectors. 2025. Online resource. Disponível em: https://github.com/sunsmarterjie/YOLOv12.

[8] LI, J. et al. Promote computer vision applications in pig farming scenarios: high-quality dataset, fundamental models, and comparable performance. Journal of Integrative Agriculture, 2024.

[9] SOUZA, J. et al. Pig aggression classification using CNN, transformers and recurrent networks. In: Anais do XIX Workshop de Visão Computacional. Rio Paranaíba/MG: [s.n.], 2024. p. 1–6.

[10] ALAMEER, A.; KYRIAZAKIS, I.; BACARDIT, J. Automated recognition of postures and drinking behaviour for the detection of compromised health in pigs. Scientific Reports, v. 10, p. 13665, 2020.

[11] PENG, N.; LI, F.; LUO, X. PLM-YOLOv5: Improved YOLOv5 for pig detection in livestock monitoring. In: 2024 International Conference on Intelligent Robotics and Automatic Control (IRAC). [S.l.: s.n.], 2024. p. 619–625.

[12] FENG, W.; WANG, K.; ZHOU, S. An efficient neural network for pig counting and localization by density map estimation. IEEE Access, v. 11, p. 81079–81091, 2023.

[13] LEE, J. et al. Sustainable self-training pig detection system with augmented single labeled target data for solving domain shift problem. Sensors, v. 25, n. 11, 2025.

[14] HE, P. et al. PDC-YOLO: A network for pig detection under complex conditions for counting purposes. Agriculture, v. 14, n. 10, 2024. ISSN 2077-0472. Disponível em: https://www.mdpi.com/2077-0472/14/10/1807.

[15] LIN, T.-Y. et al. Microsoft COCO: Common Objects in Context. 2015. Disponível em: https://arxiv.org/abs/1405.0312.