A Learnable Color Correction Matrix for RAW Reconstruction

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Liu, Anqi, Mu, Shiyi, Xu, Shugong
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913491447709696
author Liu, Anqi
Mu, Shiyi
Xu, Shugong
author_facet Liu, Anqi
Mu, Shiyi
Xu, Shugong
contents Autonomous driving algorithms usually employ sRGB images as model input due to their compatibility with the human visual system. However, visually pleasing sRGB images are possibly sub-optimal for downstream tasks when compared to RAW images. The availability of RAW images is constrained by the difficulties in collecting real-world driving data and the associated challenges of annotation. To address this limitation and support research in RAW-domain driving perception, we design a novel and ultra-lightweight RAW reconstruction method. The proposed model introduces a learnable color correction matrix (CCM), which uses only a single convolutional layer to approximate the complex inverse image signal processor (ISP). Experimental results demonstrate that simulated RAW (simRAW) images generated by our method provide performance improvements equivalent to those produced by more complex inverse ISP methods when pretraining RAW-domain object detectors, which highlights the effectiveness and practicality of our approach.
format Preprint
id arxiv_https___arxiv_org_abs_2409_02497
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle A Learnable Color Correction Matrix for RAW Reconstruction
Liu, Anqi
Mu, Shiyi
Xu, Shugong
Image and Video Processing
Computer Vision and Pattern Recognition
Autonomous driving algorithms usually employ sRGB images as model input due to their compatibility with the human visual system. However, visually pleasing sRGB images are possibly sub-optimal for downstream tasks when compared to RAW images. The availability of RAW images is constrained by the difficulties in collecting real-world driving data and the associated challenges of annotation. To address this limitation and support research in RAW-domain driving perception, we design a novel and ultra-lightweight RAW reconstruction method. The proposed model introduces a learnable color correction matrix (CCM), which uses only a single convolutional layer to approximate the complex inverse image signal processor (ISP). Experimental results demonstrate that simulated RAW (simRAW) images generated by our method provide performance improvements equivalent to those produced by more complex inverse ISP methods when pretraining RAW-domain object detectors, which highlights the effectiveness and practicality of our approach.
title A Learnable Color Correction Matrix for RAW Reconstruction
topic Image and Video Processing
Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2409.02497