Entertainment and Media - Metadata Tagging

Organizations often have large collections of digital media assets that need to have metadata associated with them in order to fully realize their value. Associating metadata with asset records makes it indexable and thus easier to categorize, search and maintain.

 

Manual curation of these assets is a labor intensive task. The collections are often large and can contain information in different formats (text, audio, visual etc.). In addition, tagging standards need to be maintained in order to ensure inconsistencies in the data are not introduced that make later retrieval of information difficult. 

Realizing the value of historical sporting digital media assets by using Artificial Intelligence (AI) to assist in doing the work of 15 interns.

AI techniques from the fields of Computer Vision, Optical Character Recognition, Speech and Natural Language Processing can be used to extract information from unstructured data in raw digital media assets. Documents can be classified into topics and named entities such as locations, events and people can be recognised.

 

The San Francisco Giants turned to Veritone’s AI-enabled digital asset management platform to help preserve 60 years of media assets nearing the end of their life. After digitizing the content, the Giants were faced with hiring 15 interns for a full year to associate metadata with the content.

 

Instead of going through the process of applying human labor to the process, the Giants used Veritone’s keyword extraction, object detection and logo detection amongst other capabilities to automatically add metadata tags instantly enabling search and discovery of the content and ultimately opportunities for monetisation.