4 ms·
It should also be possible to train it on itself to improve moving scenes by using the motion itself as temporal super-sampling, just like the human eye does.
by Mithaldu 10y ago
It should also be possible to train it on itself to improve moving scenes by using the motion itself as temporal super-sampling, just like the human eye does.
- ttul 10y agoOoh
- mm_alex 10y agothis works quite well, and does not necessarily require any NN/machine learning. see the youtube for this paper https://www.disneyresearch.com/publication/scenespace/ https://www.disneyresearch.com/publication/scenespace/ tldr simple brute force weighted average of samples from many frames, combined with a noisy/low quality depth-from-motion estimate can be used to de-noise, increase resolution and otherwise manipulate video footage. very cool paper with great results from a simple technique.