Copyright © 2026 Authors retain the copyright of this article. This article is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

@article{201001,
        author = {Shreyas Chulliyan and Samruddhi Shinde and Shlok Bam and Ishan Soman and Sujal Thakur and Asmita Patil},
        title = {WildGuard AI: A Real-Time Intelligent Poaching Detection System Using Machine Learning for Automated Monitoring and Alerts},
        journal = {International Journal of Innovative Research in Technology},
        year = {2026},
        volume = {12},
        number = {12},
        pages = {9857-9871},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=201001},
        abstract = {Poaching remains a major problem for biodiversity conservation across the globe, especially in protected forest areas where it is difficult to maintain constant monitoring of the human population. The original camera trap system did generate a lot of visual information, but due to being manually read, it created a delay in response and therefore did not stop the intruder in real time. With new advancements in computer vision, there are possibilities to use it in automating some of the duties associated with monitoring; however, most solutions currently out there provide only detection accuracy training but not actual operational alerting systems (for alerting the enforcement agent) or management of evidence (i.e., video/image record keeping) that is required for them to successfully perform their duties in the real world. This article introduces the WildGuard AI system as an automated, real-time monitoring and detection tool for poaching, which will assist enforcement agents in identifying suspicious human activity in proximity to wildlife populations and allow them to respond immediately when alerted. Our proposed system is an integrated solution that combines a lightweight deep learning object-detection model with a web-based management interface, automated email alerts, and structured evidence logging. Incoming video streams from live cameras or uploaded video files will be analyzed using an efficient inference engine to detect people (humans) and classes of animals that are relevant to the monitoring activity in real time. Each time a human is detected in the video stream (potentially a threat), evidence frames will be captured, annotated, and stored in an evidence repository and the event details recorded in an audit log for traceability. The experimental evaluations conducted show that humans were detected reliably with very low processing latencies to allow for timely delivery of alerts and to allow for continuous monitoring of threats. In addition, the results of the project support the conclusion that effective detection algorithms combined with automated alert systems and evidence management systems can close the gap between academic surveillance research and effective conservation technologies ready for deployment. The proposed framework provides the foundation for developing sustainable, effective, intelligent systems for anti-poaching surveillance in limited resource regions.},
        keywords = {Poaching Detection, Machine Learning, Real-Time Monitoring, Automated Surveillance, Real-Time Alerts.},
        month = {May},
        }