Project DetailsIn the first part of the project, our focus was on the PyNET network. This network was designed to replace the full ISP pipeline, which is responsible for the conversion of raw information detected by a digital camera sensor (known as a Bayer image or a RAW image) into the color image seen on the screen (of the DSLR camera, of the smartphone, etc.). Specifically, we tested different loss functions in order to improve PyNETs performance.
In the second part of the project, we explored additional ways to improve this performance. In addition, we experimented with two new networks which we saw as possible candidates for ISP pipeline replacement: an extremely simple CNN which won 1st place in a network design competition ran by the authors of the PyNET paper and an Autoencoder-type network taken from a previous project of our instructor (where it was used for denoising).
In terms of results, both new networks we tested achieved better performance than the Huawei ISP (mostly in terms of PSNR) but did not improve upon PyNET.
