MERIT: Multi-domain Efficient RAW Image Translation

2026-03-21

AI Summary

This work addresses the significant domain shift in RAW images captured by different camera sensors, which arises from variations in spectral response, noise characteristics, and tonal rendering, thereby hindering downstream vision tasks. To overcome this challenge, we propose the first unified multi-domain RAW image translation framework capable of achieving high-quality conversion between arbitrary camera domains using a single model. The core innovations include a sensor-aware noise alignment mechanism and a conditional multi-scale large-kernel attention module. Additionally, we introduce MDRAW, the first benchmark dataset tailored for multi-domain RAW image translation. Experimental results demonstrate that our method improves translation quality by 5.56 dB while reducing training iterations by 80%, substantially outperforming existing approaches.

Technology Category

• Machine Learning: Transfer, Domain Adaptation, Multi-Task Learning
• Computer Vision: Multi-modal Vision
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Security and Privacy: Data transparency and provenance
• Responsible Web: Algorithmic accountability and transparency on the web

Abstract

RAW images captured by different camera sensors exhibit substantial domain shifts due to varying spectral responses, noise characteristics, and tone behaviors, complicating their direct use in downstream computer vision tasks. Prior methods address this problem by training domain-specific RAW-to-RAW translators for each source-target pair, but such approaches do not scale to real-world scenarios involving multiple types of commercial cameras. In this work, we introduce MERIT, the first unified framework for multi-domain RAW image translation, which leverages a single model to perform translations across arbitrary camera domains. To address domain-specific noise discrepancies, we propose a sensor-aware noise modeling loss that explicitly aligns the signal-dependent noise statistics of the generated images with those of the target domain. We further enhance the generator with a conditional multi-scale large kernel attention module for improved context and sensor-aware feature modeling. To facilitate standardized evaluation, we introduce MDRAW, the first dataset tailored for multi-domain RAW image translation, comprising both paired and unpaired RAW captures from five diverse camera sensors across a wide range of scenes. Extensive experiments demonstrate that MERIT outperforms prior models in both quality (5.56 dB improvement) and scalability (80% reduction in training iterations).

Problem

• RAW image translation
• domain shift
• multi-domain
• sensor-specific noise
• computer vision

Innovation

• multi-domain RAW translation
• sensor-aware noise modeling
• unified image translation framework
• large kernel attention
• MDRAW dataset