Efficient AI for Wildlife Conservation - Challenges and Solutions for Environmental Monitoring

Watch a technical talk exploring the challenges and opportunities of applying AI to wildlife conservation monitoring. Learn how ecological data collection in the field faces unique obstacles including spatiotemporal correlations, imperfect data quality, and fine-grained categories that current deep learning methods struggle to handle. Discover the constraints of implementing AI systems in remote areas with limited bandwidth, power, storage and computational resources. Gain insights into open research problems where developing more robust, efficient and adaptable AI models could significantly impact environmental monitoring and conservation efforts. The speaker, Sara M. Beery, a Visiting Researcher at Google and Assistant Professor at MIT CSAIL, discusses how automated data processing solutions are essential for scaling up wildlife monitoring to understand and improve conservation outcomes in real-time.