# W-Net: Two-stage U-Net with misaligned data for raw-to-RGB mapping

arXiv: https://arxiv.org/abs/1911.08656 · HTML: https://mlsanity.com/paper/1911.08656/

## Key facts

- Cited by 7 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 20 Nov 2019.

## Built on by (most cited first, 7 of 7)

- [PyNET-CA: Enhanced PyNET with Channel Attention for End-to-End Mobile Image Signal Processing](https://mlsanity.com/paper/2104.02895/) (2021)
- [PyNET-QxQ: An Efficient PyNET Variant for QxQ Bayer Pattern Demosaicing in CMOS Image Sensors](https://mlsanity.com/paper/2203.04314/) (2022)
- [Rendering Nighttime Image Via Cascaded Color and Brightness Compensation](https://mlsanity.com/paper/2204.08970/) (2022)
- [ISP meets Deep Learning: A Survey on Deep Learning Methods for Image Signal Processing](https://mlsanity.com/paper/2305.11994/) (2023)
- [RPBA-Net: An Interpretable Residual Pyramid Bilateral Affine Network for RAW-Domain ISP Enhancement](https://mlsanity.com/paper/2605.03626/) (2026)
- [Lightweight Unpaired Smartphone ISP Transfer with Semantic Pseudo-Pairing](https://mlsanity.com/paper/2605.07495/) (2026)
- [Visual Autoregressive Priors for RAW-to-sRGB Image Signal Processing](https://mlsanity.com/paper/2609.18302/) (2026)

## Abstract (arXiv)

Recent research on learning a mapping between raw Bayer images and RGB images has progressed with the development of deep convolutional neural networks. A challenging data set namely the Zurich Raw-to-RGB data set (ZRR) has been released in the AIM 2019 raw-to-RGB mapping challenge. In ZRR, input raw and target RGB images are captured by two different cameras and thus not perfectly aligned. Moreover, camera metadata such as white balance gains and color correction matrix are not provided, which makes the challenge more difficult. In this paper, we explore an effective network structure and a loss function to address these issues. We exploit a two-stage U-Net architecture and also introduce a loss function that is less variant to alignment and more sensitive to color differences. In addition, we show an ensemble of networks trained with different loss functions can bring a significant performance gain. We demonstrate the superiority of our method by achieving the highest score in terms of both the peak signal-to-noise ratio and the structural similarity and obtaining the second-best mean-opinion-score in the challenge.