Improving asset find-ability with AI metadata generation

• Background

  A world leader in file digitization and restoration of valuable archives came to Clarifai to improve the metadata tagging. The company specializes in transferring old film archives into digital files. The company’s software for “smart indexing” wanted to use AI and machine learning to dramatically increase the speed and precision with which metadata can be included within the film asset.

• Information

  Use Case

  Metadata generation

  Industry

  Media and entertainment

  Client

  Media and Digital Advertising Agency

Challenge

A period of investigating the possible alternative approaches led the company to Clarifai's AI driven computer vision platform and its ability to automate the tagging process. Using Clarifai, the company was able to automatically detect objects that appeared in the archived material with the highest accuracy and faster than a human eye could ever achieve. This meant that the creation of easily searchable metadata was transformed as was the value of the archived asset.

Solution

A period of investigating the possible alternative approaches led the company to Clarifai.  Clarifai's AI driven computer vision platform could take the place of a human viewing all the footage in its entirety while frantically taking notes. Using Clarifai, the company was able to automatically detect objects that appeared in the archived material with the highest accuracy and faster than a human eye could ever achieve. This meant that the creation of easily searchable metadata was transformed as was the value of the archived asset.

Results

With a team of two talented developers, the company built a platform that automatically created a keyframe every third second, while digitizing old film content. After which it sent the image to the Clarifai API – returning it, fully indexed, in real time.

• Clarifai’s computer vision AI can understand concepts with better-than-human accuracy in many cases and certainly faster than the human eye could achieve. The creation of easily searchable metadata has made the value of our archived assets even greater to our customers."

  CEO & Founder