Toshiba AI Technology Catalog

  • Media data analysis

Deep learning image quality improvement

Accurately remove noise in adapting to noise level of input images.


  • Neural network activation function proposed by Toshiba achieves both high performance noise removal and ease of use to enable simple adjustments in adapting to noise level of input images.

Applications



  • Improve quality of images from on-board and security cameras.
  • Improve quality of images from various image sensors.

Benchmarks, strengths, and track record



  • In the case of images with unexpected noise level outside of the scope of prior learning, improved restoration error to less than one third of conventional deep learning noise removal technologies.

Inquiries



Please include the title “Toshiba AI Technology Catalog: Deep learning image quality improvement” or the URL in the inquiry text.
Please note that because this technology is currently the subject of R&D activities, immediate responses to inquiries may not be possible.

References: