Making high-quality, annotated wildlife video data and powerful AI models publicly available lowers barriers for conservation analytics, enabling faster detection of population declines, creation of biodiversity credits, and better-informed policy and resource allocation. This could materially improve conservation outcomes by scaling monitoring and turning observational data into actionable insights.
Meta and Conservation X Labs are deploying advanced AI — including SAM 3 and CM3 — to automate identification and behavioral monitoring of wildlife in camera-trap videos, enabling precise individual-level tracking rather than simple bounding boxes. The partners will release an open dataset of 12,000 annotated videos covering more than 100 species to accelerate model development and benchmarking. This combination of foundational vision models and public data promises scalable, cross-environment tools for tracking populations and detecting changes linked to threats like climate change. Conservation X Labs says the technology will dramatically improve accuracy and speed of conservation monitoring and research.
Comments
Want to join the conversation?
Loading comments...