Closed KatsarosEf closed 3 years ago
Thank you!
Regarding the Poisson noise, I haven't performed experiments with this model in particular, but I have done with other similar CNN denoisers. It is important to update the noise map accordingly to reflect the distribution of the new type of noise. After this, the model should be retrained with the new noise distribution.
As for the second point, I expect the performance is somehow inferior with respect to the version with 5 input frames, but in any case it should do a good job.
Good luck with your experiments and keep me posted on your results. Let me know if you have further questions.
Congratulations on your paper, really nice and well-explained work! I am implementing some modifications, namely, I am adding extra Poisson noise on the videos. You mention that the method can be extended to Poisson, but did you perform any such experiments of your own? Had you identified potential pitfalls?
Moreover, I am adjusting your method to work independently of future frames. That said, I am employing frames t-3, t-2, t-1 and t in 2 triplet combinations [(t-2, t-1, t), (t-3, t-1, t)] utilizing two blocks instead of three. You have performed a relevant ablation study (one block) which was explanatory, thanks!