Deep Render - innovation in image compression tech

by Startacus Admin

London-based startup Deep Render tackles the internet’s data and bandwidth challenges via its innovative AI-powered file compression technology.

The internet is one of the biggest inventions in human history. It allows us to learn beyond school, it allows us to connect with people all around the world and tell them their opinions are wrong, it lets us stay informed (theoretically, ignoring contrary evidence), and it gives us access to more entertainment. The more technology advances, the more information can be pushed through the internet tubes into our homes; however, this means larger quantities of data, ideally at faster and faster speeds.

This is where compression comes in. Compression does what it sounds like: it compresses the data to a smaller size (you don’t want it getting stuck in those tubes and causing a clog), and then when you receive it, it’s restored. The problem is that the technology to do this is outdated and isn’t evolving at nearly the speed it needs to to keep up. We can see when things haven’t gone well when we try to stream a video or film and it appears pixelated and contains artefacts. Rather than ISPs spending billions on constantly improving bandwidth, perhaps there’s another solution.

London-based startup Deep Render is taking on the challenge of bringing modern compression up to speed. The startup says that the current progression of lossless compression is only a 15-20% improvement every 8-10 years; using AI to change the fundamental way that compression works, Deep Render can immediately provide a 500% improvement, with an additional improvement of over 20% every year. Using a neural network for compression, the AI-codec optimises data quality for human perception rather than the traditional hard figures demanded by hardware, which also means that it doesn’t need to rely on hardware-accelerated and instead is software-driven.

The startup says that their technology is already 80% better than the most state-of-the-art traditional compression codec. And 80% better means a significant reduction in costs for not only ISPs but for the likes of Netflix, whom the startup estimates could save around $1.6 billion per year on delivery costs by using their AI-codec. 

In March, Deep Render announced that they had raised $9 million in a Series A round led by IP Group and Pentech Ventures, and including a competitive grant of $2.7 million from the European Innovation Council. This funding will allow the startup to scale up their staff numbers and expand in the US, as well as garnering more commercial engagement. The tech is already being trialled by select companies, and with the above figures, it’s no wonder they are calling this the ‘iPhone moment for the compression industry’.

 

 

Subscribe to our newsletter

If you would like to receive our startup themed newsletter, full of the latest startup opportunities, events, news, stories, tips and advice, then sign up here.

• #london
• #startup
• #innovation
• #compression
• #ai
• #machine learning

Published on: 17th May 2023