With the help of Twitter’s Cortex, Periscope could soon be able to scan and identify the content within live streams. Twitter have created a new system using a deep learning algorithm which allows it to categorise a selection of different streams. Intended to help deliver an increasingly seamless and advanced discovery experience for users, the new technology could potentially also help to rid the network of unsavoury content…
Twitter’s Cortex is a team of engineers, data scientists and machine learning researchers who focus on developing ease of discovery and systems which reflect the users and content on their products. They first unveiled this new live streaming scanning technology to MIT Technology Review, managing to successfully scan and categorise 24 live streams.
Over recent years researchers have been successful in developing algorithms that identify and categorise content within stationary images and photographs. Twitter’s Cortex are the first to have substantially made progress in the development of products that can do this for live videos. Live streams are particularly tricky due to the constantly altering nature of the content within them and also the variance in video quality. This is how MIT Technology Review described the technology:
“Twitter effectively built a custom supercomputer made entirely of graphics processing units (GPUs) to perform the video classification and serve up the results. These chips are especially efficient for the mathematical calculations required for deep learning, but normally they are just on part of a larger computer system… The Cortex team has ambitions to develop a sophisticated recommendation system to help filter and curate all sorts of content shared through the service, based on a user’s previous activity.”
What could the new scanning algorithm be used for?
The team’s primary intentions for the algorithm is to help develop content discovery on the network, such as search within the app. Although the project is in it’s early stages, live streaming is becoming increasingly prominent in the world of social networking, and so the development of search and discovery tools such as these will eventually become incredibly important. The algorithm could also prove very useful for tackling the issue of “dark” content in the network, something that has been all over media recently.
