Cloud

New machine learning models boost video and language capabilities for Google Cloud users

Google Cloud's Cloud Video Intelligence and Cloud Natural Language machine learning models are now generally available. Here's what they enable for the enterprise.

Google Cloud launched two of its machine learning models in general availability on Tuesday, helping customers more easily analyze video content and classify content by language.

The two models—Cloud Video Intelligence and Cloud Natural Language—were both detailed in a Google press release.

The Cloud Video Intelligence API was initially launched earlier in 2017. Essentially, this API analyzes video content for specific entities, and makes them searchable for users. Working with beta users, the release said, Google has been able to improve the product's accuracy and add new features.

SEE: Job description: Cloud engineer (Tech Pro Research)

Entities in a video are basically just any noun that may appear. This could be the presence of a dog, car, or flower, for example. The API can also determine when a scene has changed in a video.

According to the release, the tool can now detect 20,000 labels, including 667 car models and more than 200 types of buildings, and can be used to moderate content as well. A demo video is available on YouTube, and the code can be found on GitHub.

Additionally, a new video transcription service for Cloud Video Intelligence is also available in private beta. This feature allows users to "automatically transcribe video audio into text," the release said.

Another generally available feature is Content Classification for Cloud Natural Language. The Cloud Natural Language API analyzes the structure and meaning of text through a REST API. The new Content Classification feature automatically sorts text documents into specific categories like health or news, the release said. Currently, there are more than 700 categories available.

One example for the real-world application of Content Classification is for media companies that need to understand the business value of their content more fully, the release said. "This means they can build or enhance recommended content for their users as well as gain a new layer of data to help guide opportunities for their advertisers," the release said.

Interested users can get started with Content Classification by reading Google's tutorial here.

The 3 big takeaways for TechRepublic readers

  1. Google Cloud launched two new machine learning models into general availability—Cloud Video Intelligence and Cloud Natural Language.
  2. Cloud Video Intelligence detects the presence of entities in video as well as scene changes, and can now transcribe video into text.
  3. Cloud Natural Language has a new feature called Content Classification that sorts text documents into one of more than 700 available categories, like government or health.

Also see

cloud.jpg
Image: iStockphoto/chombosan

About Conner Forrest

Conner Forrest is a Senior Editor for TechRepublic. He covers enterprise technology and is interested in the convergence of tech and culture.

Editor's Picks

Free Newsletters, In your Inbox