GrayMeta unlocks hidden value stored in any asset by leveraging AI and machine learning
Caspar Herzberg |
This article first appeared in the
Winter issue of The Record.
Companies’ digital files are valuable assets, but many organisations are drowning in a sea of their own data, where they struggle to find content they know they own.
GrayMeta was founded to create a lifeline for those companies. Its flagship product, the GrayMeta Platform, connects to a customer’s storage – on-premise or in the cloud – extracts technical metadata from all assets, creates temporal metadata using AI and machine learning, and then still allows users to curate that metadata if needed. With major motion picture studios and broadcasters among its customers, GrayMeta wanted to go beyond the technical metadata and provide insight into the content of videos and images, making it easily searchable.
GrayMeta considered using open-source solutions to analyse the content of video and photo files and collect useful data about what they contain. However, these required extensive training and contained numerous errors. After learning about the Microsoft Cognitive Services APIs, the company chose to incorporate the Computer Vision API, Face API and Emotion API into the GrayMeta Platform, to sample frames from videos and send them to the APIs for analysis.
Key features of the Computer Vision API include facial recognition, celebrity detection, image recognition, adult content detection, and the fact that the tool is pretrained. The Emotion API is used to analyse faces and determine what sort of emotion the person is displaying. Combining the two APIs makes it possible to search for a video of a specific celebrity looking happy, sad or surprised, for example.
Integrating the APIs into the GrayMeta Platform was easy. While it took a few months to build the programmatic infrastructure to manipulate all the tagging data, the actual process of implementing the APIs took only a few hours. This speed is critical for agile development, enabling the team to iterate quickly and to easily incorporate a ¬customer-driven concept like “I need to be able to search for a celebrity” into its product.
With the ongoing advancement of their solution, GrayMeta has found success in the broadcast and media and entertainment industries in the US and UK, and has most recently expanded their reach to Asia and Australia. “Because the industry is producing content at an unbelievable rate, we’ve focused on solving media workflow inefficiencies,” says Tom Szabo, Chief Executive Officer of GrayMeta. “By leveraging Microsoft Cognitive Services in the platform, our clients are finding that they can save time and money on content creation, management, and even delivery. Our goal is that they no longer struggle to find what they own, but will instead gain more insight from it.”