Cisco Blogs
Share

Powering AI Workloads in Media and Entertainment


September 14, 2018 - 0 Comments

If you’ve been following news in Artificial Intelligence (AI) and are involved in media and entertainment, this has been a week of milestones.

First, Cisco announced the expansion of the UCS portfolio to power the full lifecycle of AI, from data collection and analysis, to data preparation and training, to the real-time inference.

And this week at the #IBC2018 trade show in Amsterdam, with our partners NVIDIA, we are showcasing how GPU-accelerated processing of data can transform how content is created and curated.

The power of the possible is visible on our booth. Here are some use cases NVIDIA are investigating:

  • Imagine being able to use AI to avoid re-shooting a scene, with tools that can enable a director or producer to insert or remove objects based on analysis, training and inference of data. Noisy images can similarly be cleaned up.
  • Say you are shooting a scene at 60 frames per second (FPS), but you want to create a really cool slow-motion effect. AI can help you to create hundreds of additional frames per second allowing you to interpolate what was in the frames before and after an actual shot.
  • Or come learn about GANS (generative adversarial networks) which consists of two neural networks where AI can enable images to be inserted into a framework based on machine learning/training of images.

The possibilities are endless. And the foundation is a data center infrastructure that powers the massive collection, analysis and training of data.

See it in our #CiscoIBC booth: September 14-18. Stand 1.A71

Using AI to Avoid Reshoots

Using AI to Create Additional Frames per Second

 

 

 

 

 

 

 

 

 

 

 

Leave a comment

We'd love to hear from you! Your comment(s) will appear instantly on the live site. Spam, promotional and derogatory comments will be removed and HTML formatting will not appear.