There is poor quality video, poorly visible objects. The task is to recognize objects in the video. Who faced video enhancement?

I think this is quite realistic with the use of neural networks.

  • 2
    Neural networks will simply think out objects based on their experience - andreymal
  • Yes, I understand that ... - Igor Lavrynenko

1 answer 1

There are developments for the photo, example:

enter image description here

with only 340 lines of code

There are developments for poorer quality, example:

enter image description here

But keep in mind that in any case, the worse the quality of the original image, the more improbable the recovery.

These developments can be altered for video, but it already depends on your resources.