Recent Advances in AI-Based Wildlife Monitoring: A Review of Multimodal Detection, Edge Computing, and Conservation Applications

Recent advancements in the field of artificial intelligence-based wildlife monitoring systems, based on image-based detection, bioacoustic analysis, IoT-enabled alert system, and energy efficient edge deployment, were discussed in this review. The findings demonstrate that deep learning models, specifically CNN and YOLO, are very accurate on visual detection tasks, but transformer-based models, such as Wav2Vec2, were much better on acoustic recognition. The approach of transfer learning also allows recognizing a particular animal in detail, which demonstrates that it is possible to conduct in-depth ecological monitoring. Despite these advancements, none of the detection modalities can be applied to consistently achieve good results in all environmental conditions. Visual systems are influenced by lighting, occlusion and camouflage, and acoustic systems are influenced by environmental noise and redundant signals. Multimodal methods, which are based on visual, acoustic and sensor-based information, have proven to be the most effective approach in improving reliability and robustness in real life scenarios. System design by energy-efficient design, especially through edge computing and motion-based filtering techniques is a key focus enabling long term deployment in remote environments. These strategies reduce the amount of computation required, and dramatically extend the operational life, allowing large-scale monitoring to become more feasible. Overall, the field of wildlife monitoring has been revolutionized by artificial intelligence through offering automated, real-time and scalable solutions. Multi-modal integration, model design and affordability deployment strategies should be prioritized in the future to facilitate sustainable conservation measures and eliminate human-wildlife conflicts.

1.      Norouzzadeh, M.S.; Nguyen, A.; Kosmala, M.; Swanson, A.; Palmer, M.S.; Packer, C.; Clune, J. Automatically identifying, counting, and describing wild animals in camera-trap images with deep learning. Proc. Natl. Acad. Sci. USA 2018, 115, E5716–E5725. https://doi.org/10.1073/pnas.1719367115 2.      Premarathna, K.S.P.; Rathnayaka, R.M.K.T.; Charles, J. To solve the issue of human-elephant conflict and to track the movements of an elephant with the help of deep learning, an elephant detection system. In Proceedings of the IEEE International Conference on Industrial and Information Systems (ICIIS), Colombo, Sri Lanka, 2020. https://doi.org/10.1109/ICITR51448.2020.9310798 3.      Sato, R.; Saito, H.; Tomioka, Y.; Kohira, Y.The energy reduction method in detecting devices of wild animals. IEEE Access 2022, 10, 24149–24160. https://doi.org/10.1109/ACCESS.2022.3155242 4.      Redmon, J.; Farhadi, A. YOLOv3: A slight improvement. arXiv 2018, arXiv:1804.02767. 5.      Jocher, G.; Chaurasia, A.; Qiu, J. YOLOv5 by Ultralytics. Available online: https://github.com/ultralytics/yolov5 (accessed on 10 April 2026). Jocher, G.; Qiu, J.; Chaurasia, A. YOLOv8: Ultralytics next-generation object detection model. Available online: https://github.com/ultralytics/ultralytics (accessed on 10 April 2026). 6.      Tan, M.; Le, Q. EfficientNet: Rethinking model scaling for convolutional neural networks. This can be found in Proceedings of the International Conference on Machine Learning (ICML), Long Beach, CA, USA, 2019. https://doi.org/10.48550/arXiv.1905.11946 7.      Baevski, A.; Zhou, H.; Mohamed, A.; Auli, M. Wav2Vec 2.0: A framework for self-supervised learning of speech representations. In Proceedings of the Advances in Neural Information Processing Systems (NeurIPS), Vancouver, Canada, 2020. https://doi.org/10.48550/arXiv.2006.11477 8.      Chollet, F. Deep learning with Python; Manning Publications: New York, NY, USA, 2017. 10. 9.       He, K.; Zhang, X.; Ren, S.; Sun, J. Deep residual learning for image recognition. Published in proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Las Vegas, NV, USA, 2016. https://doi.org/10.1109/CVPR.2016.90 10.   Very convolutional networks, very large-scale image recognition. Simonyan, K.; Zisserman, A. arXiv 2015, arXiv:1409.1556.