Accessing Video Footage: Saint Louis Zoo

The Arizona State University Artificial Intelligence Cloud Innovation Center (AI CIC), powered by Amazon Web Services (AWS), collaborated with the Saint Louis Zoo to develop an AI-integrated video cataloging and search platform. The goal of this project is to make the Zoo’s extensive video archive of animal behaviors more accessible and valuable to researchers, educators, and conservationists. Over the years, the Saint Louis Zoo has collected a vast amount of video footage that provides critical insights into animal behavior, but much of these data remain underutilized due to the lack of an organized, searchable system. To address this challenge, the project aims to automate video tagging, enhance search capabilities, and generate short previews to streamline video discovery and accessibility.

The Zoo has accumulated thousands of hours of video footage documenting animal behaviors, but the lack of a structured video management system makes it difficult for researchers and the public to access relevant footage. Currently, videos are not tagged or categorized efficiently, making it time-consuming to locate specific content. Researchers and educators need a smart search system that allows them to filter videos by species, habitat, and other key metadata. Additionally, manually reviewing full-length videos is impractical, so short previews are necessary to help users quickly determine video relevance before requesting full access. Without a robust search and tagging solution, valuable research material remains inaccessible and difficult to utilize.

The Zoo video project introduces a web-based platform that enables users to search, preview, and request video footage using an AI-powered tagging system. The Admin Tool allows Zoo administrators to upload videos, assign metadata tags, and manage the library, categorizing content by species, location, and other variables for quick retrieval. The User-Facing Web Portal lets researchers and educators search videos, view 5-second previews, and request full-length downloads with a shopping cart feature.

Built on AWS cloud services, the system uses Amazon S3 for storage, AWS Lambda for video processing, and Amazon Bedrock (Claude 3.5 Sonnet) for AI-driven metadata extraction. Amazon OpenSearch powers fast, accurate searches, while Amazon ECS with Fargate handles video clipping and thumbnail generation. AWS Cognito ensures secure authentication, AWS Step Functions manage workflows, and AWS SES sends email notifications for access requests.

By automating video tagging and search, this system makes animal research footage more accessible, searchable, and efficiently organized.

The ZVDB system is setting a new standard for efficiently managing and accessing large-scale video archives. Beyond zoos, this AI-powered video search and tagging system offers valuable applications across various industries. In wildlife conservation and research, organizations studying animal behavior can leverage AI-powered tagging to quickly retrieve data and conduct behavioral analysis. Aquariums and marine research centers can use this technology to catalog and analyze marine life footage, aiding in conservation efforts. 

Educational institutions and museums benefit by providing students and researchers with structured access to archival or sports footage, enhancing learning and entertainment experiences. Additionally, media and documentary producers can efficiently search and retrieve high-quality footage for educational and commercial content. By harnessing AI-driven automation for video cataloging, this solution has the potential to revolutionize digital asset management, ensuring that valuable footage is accessible, searchable, and beneficial for scientific, educational, and creative advancements.