Reports from the Field
Deep machine learning applied to moving image restoration
Lessons learned over seven years

We started restoring multispectral moving images using deep machine learning in 2017. This time we would like to share our expertise and experiences specifically in the field of machine learning applied to restoration. There is another meeting focussing on Working with multispectral moving images.

What and how

Theoretical introduction to supervised training of artificial neural networks and case studies presented by Michal Cohen and Reto Kromer, followed by an informal Q&A session and light refreshments. The participants should be familiar with the basics of the so-called artificial intelligence.

When and where

In-person at the EPFL Innovation Park with webcast on Friday the 15th of March 2024, starting at 11 o’clock sharp here in Switzerland (that’s 10:00 UTC). We will use a Jitsi-based platform and the link will be provided in proper time.

As usual, the admission to our Reports from the Field is free, but a registration is required, as the seats are limited to a dozen or so, in order to enable in-depth technical discussions and the exchange of information.


  • We are very unhappy that our generosity is often betrayed, as many of our materials are distributed without naming their origin. Increasingly, others are even shamelessly claiming authorship of our work, which is just disgusting behaviour. And that’s the very reason why some resources are no longer freely available, but have been moved to password-protected parts of our website.