# R3: Reconstruction, Raw, and Rain: Deraining Directly in the Bayer Domain

arXiv: https://arxiv.org/abs/2509.24022 · HTML: https://mlsanity.com/paper/2509.24022/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 22 of its 28 references that are in the map.
- First posted on arXiv 28 Sep 2025.

## Builds on (most cited first, 22 of 22)

- [U-Net: Convolutional Networks for Biomedical Image Segmentation](https://mlsanity.com/paper/1505.04597/) (2015)
- [CBAM: Convolutional Block Attention Module](https://mlsanity.com/paper/1807.06521/) (2018)
- [Benchmarking Robustness in Object Detection: Autonomous Driving when Winter is Coming](https://mlsanity.com/paper/1907.07484/) (2019)
- [Density-aware Single Image De-raining using a Multi-stream Dense Network](https://mlsanity.com/paper/1802.07412/) (2018)
- [Recurrent Squeeze-and-Excitation Context Aggregation Net for Single Image Deraining](https://mlsanity.com/paper/1807.05698/) (2018)
- [Spatial Attentive Single-Image Deraining with a High Quality Real Rain Dataset](https://mlsanity.com/paper/1904.01538/) (2019)
- [DeepISP: Towards Learning an End-to-End Image Processing Pipeline](https://mlsanity.com/paper/1801.06724/) (2018)
- [Syn2Real Transfer Learning for Image Deraining using Gaussian Processes](https://mlsanity.com/paper/2006.05580/) (2020)
- [Learning Raw Image Denoising with Bayer Pattern Unification and Bayer Preserving Augmentation](https://mlsanity.com/paper/1904.12945/) (2019)
- [Beyond Monocular Deraining: Parallel Stereo Deraining Network Via Semantic Prior](https://mlsanity.com/paper/2105.03830/) (2021)
- [Semi-Supervised Semantic Segmentation via Marginal Contextual Information](https://mlsanity.com/paper/2308.13900/) (2023)
- [ISP Distillation](https://mlsanity.com/paper/2101.10203/) (2021)
- [Dual-Camera Joint Deblurring-Denoising](https://mlsanity.com/paper/2309.08826/) (2023)
- [Noisy Annotations in Semantic Segmentation](https://mlsanity.com/paper/2406.10891/) (2024)
- [Raw Bayer Pattern Image Synthesis for Computer Vision-oriented Image Signal Processing Pipeline Design](https://mlsanity.com/paper/2110.12823/) (2021)
- [Gradient-based Feature Extraction From Raw Bayer Pattern Images](https://mlsanity.com/paper/2004.02429/) (2020)
- [AMED: Automatic Mixed-Precision Quantization for Edge Devices](https://mlsanity.com/paper/2205.15437/) (2022)
- [WAVECLIP: Wavelet Tokenization for Adaptive-Resolution CLIP](https://mlsanity.com/paper/2509.21153/) (2025)
- [Stereo Image Rain Removal via Dual-View Mutual Attention](https://mlsanity.com/paper/2211.10104/) (2022)
- [Multi-dimension Queried and Interacting Network for Stereo Image Deraining](https://mlsanity.com/paper/2309.10319/) (2023)
- [Hysteresis Activation Function for Efficient Inference](https://mlsanity.com/paper/2411.10573/) (2024)
- [DiffStereo: High-Frequency Aware Diffusion Model for Stereo Image Restoration](https://mlsanity.com/paper/2501.10325/) (2025)

## Abstract (arXiv)

Image reconstruction from corrupted images is crucial across many domains. Most reconstruction networks are trained on post-ISP sRGB images, even though the image-signal-processing pipeline irreversibly mixes colors, clips dynamic range, and blurs fine detail. This paper uses the rain degradation problem as a use case to show that these losses are avoidable, and demonstrates that learning directly on raw Bayer mosaics yields superior reconstructions. To substantiate the claim, we (i) evaluate post-ISP and Bayer reconstruction pipelines, (ii) curate Raw-Rain, the first public benchmark of real rainy scenes captured in both 12-bit Bayer and bit-depth-matched sRGB, and (iii) introduce Information Conservation Score (ICS), a color-invariant metric that aligns more closely with human opinion than PSNR or SSIM. On the test split, our raw-domain model improves sRGB results by up to +0.99 dB PSNR and +1.2% ICS, while running faster with half of the GFLOPs. The results advocate an ISP-last paradigm for low-level vision and open the door to end-to-end learnable camera pipelines.