Tuesday, September 25, 2018

Fujitsu Presents 1,000:1 Image Compression for Driverless Cars

Nikkei: Fujitsu unveils an image compression technology that reduce the file size to 0.1% of the original one, aimed to self-driving applications.

The new algorithm is said to compress data with 1,000:1 ratio, while still maintaining its usability for AI ML algorithms. By reducing the volume of data, cars can send high-resolution images over the internet to data centers in real time.

1 comment:

  1. "The new algorithm, which draws on Fujitsu's know-how in 4K and 8K video broadcasts, compresses data at a ratio of 1,000-to-1 compared to the more typical 80-to-1. By significantly reducing the volume of data, moving cars will be able to send high-resolution images over the internet to data centers in real time"

    I'm very suspicious they can claim such compression ratio regardless of the entropy of the image-under-consideration...

    ReplyDelete

All comments are moderated to avoid spam and personal attacks.