This is a result of improving the trainable parameter requirement from O(N4) to O(m), where N is pixel size and m is numberoffibremodes.

We develop a new type of model for solving the task of inverting the transmission effects of multi-mode optical fibres through the construction of an $\mathrm{SO}^{+}(2,1)$-equivariant neural network. This model takes advantage of the of the azimuthal correlations known to exist in fibre speckle patterns and naturally accounts for the difference in spatial arrangement between input and speckle patterns. In addition, we use a second post-processing network to remove circular artifacts, fill gaps, and sharpen the images, which is required due to the nature of optical fibre transmission. This two stage approach allows for the inspection of the predicted images produced by the more robust physically motivated equivariant model, which could be useful in a safety-critical application, or by the output of both models, which produces high quality images. Further, this model can scale to previously unachievable resolutions of imaging with multi-mode optical fibres and is demonstrated on $256 \times 256$ pixel images. This is a result of improving the trainable parameter requirement from $\mathcal{O}(N^4)$ to $\mathcal{O}(m)$, where $N$ is pixel size and $m$ is number of fibre modes.

We develop a new type of model for solving the task of inverting the transmission effects of multi-mode optical fibres through the construction of an \mathrm{SO} { }(2,1) -equivariant neural network. This model takes advantage of the of the azimuthal correlations known to exist in fibre speckle patterns and naturally accounts for the difference in spatial arrangement between input and speckle patterns. In addition, we use a second post-processing network to remove circular artifacts, fill gaps, and sharpen the images, which is required due to the nature of optical fibre transmission. This two stage approach allows for the inspection of the predicted images produced by the more robust physically motivated equivariant model, which could be useful in a safety-critical application, or by the output of both models, which produces high quality images. Further, this model can scale to previously unachievable resolutions of imaging with multi-mode optical fibres and is demonstrated on 256 \times 256 pixel images.

I have read this paper and found it to be a solid contribution both to the field of optics and to NIPS. With respect to its contribution to optics, I will admit that I have not worked in this field for over two decades, but it does seem that they are solving an interesting problem in a new way. Most cited (2010) "General Bayesian estimation for speckle noise reduction in optical coherence tomography retinal imagery" Alexander Wong, Akshaya Mishra, Kostadinka Bizheva, David A. Clausi1 And I could not find a similar work in my search. The authors contribute compelling examples in their supplemental materials, which lends credence to the claims that their method actually works. The authors also promise to contribute a novel dataset to use for ML benchmarking on this type of problem.