PMY & Google Cloud Partnership - Video Intelligence
SpringHill Migration to Google Cloud for Video Intelligence
PMY used Google Cloud's Video Intelligence API to build an automated video processing tool. Using restored media from off-site backups, the solution integrated AI-enhanced metadata from Google’s Video Intelligence API.
The challenge
Over 700TB of video was restored from off-site backups efficiently and accurately, to generate usable and valuable metadata. The metadata was then pushed back into Iconik, SpringHill’s Cloud Media Asset Management System. Because of file size and run time limitations, PMY developed special processes to chunk larger videos into processable sizes and recombine the generated pieces metadata to fit the original video.
The solution
PMY used Google Cloud's Video Intelligence API to build an automated video processing tool. Using restored media from off-site backups, the solution integrated AI-enhanced metadata from Google’s Video Intelligence API to automatically tag entire videos and ingest their metadata into Iconik, SpringHill’s Cloud Media Asset Management System.
The result
Transferred 700TB of data into Google Cloud Storage from off-site/physical locations. A selection of this data was processed through the Video Intelligence API via Cloud Run jobs. BigQuery was used as a control table to prevent duplicate processing of files as well as hold the generated metadata from the Video Intelligence API. The metadata was pushed to Iconik by a Cloud Scheduler job.
The result:
Rich, searchable metadata inside SpringHill's Media Asset Management System to drive personalized user content experiences, enhance content creation, and provide analytics about view preferences. A pipeline for consumption of future SpringHill video and resulting metadata.
Delivered future roadmap for generating metadata for live streaming video as well as performing advance.