New AI System Enhances Wildlife Conservation Efforts by Tracking Nearly 100 Species
Visual AI tracks nearly 100 wildlife species to improve conservation
Image: Phys.org
A new AI system, SA-FARI, developed by a University of Bristol-led team, can automatically identify and track nearly 100 wildlife species in video footage. This innovation aims to significantly reduce the time researchers spend analyzing wildlife data, enhancing conservation efforts worldwide.
- 01SA-FARI utilizes META's Segment Anything Model 3 (SAM3) to track individual animals in videos.
- 02The project compiled over 11,000 annotated wildlife videos to train the AI system.
- 03SA-FARI has been nominated for an award at the Conference for Computer Vision and Pattern Recognition (CVPR).
- 04The AI system can accurately separate animals from their backgrounds, aiding in behavioral analysis.
- 05Future enhancements may include tracking animal body poses and providing natural language descriptions.
Advertisement
In-Article Ad
The SA-FARI project, led by a team from the University of Bristol in collaboration with ConservationX Labs (CXL) and META, introduces an advanced AI system capable of tracking nearly 100 wildlife species in video footage. This system employs META's Segment Anything Model 3 (SAM3) to automatically identify and follow individual animals, which allows researchers to analyze wildlife behavior more effectively. By utilizing a dataset of over 11,000 annotated wildlife videos, the SA-FARI project aims to save researchers thousands of hours previously spent on manual analysis. The project will be presented at the Conference for Computer Vision and Pattern Recognition (CVPR) in Denver, where it has been nominated for an award. According to Professor Tilo Burghardt from the University of Bristol, this initiative exemplifies the importance of global solutions for wildlife conservation. The freely available dataset is expected to enhance ecological research and conservation efforts worldwide, with potential future developments including tracking animal body poses and offering natural language descriptions.
Advertisement
In-Article Ad
The SA-FARI project is expected to streamline wildlife monitoring efforts, which is crucial for effective conservation strategies.
Advertisement
In-Article Ad
Reader Poll
How do you feel about the use of AI in wildlife conservation?
Connecting to poll...
Read the original article
Visit the source for the complete story.