2019 · 10 citations · 16 references
ComputingComputer ScienceComputer VisionMachine LearningMachine VisionColor ConstancyColor ImagingImage ReconstructionImage SynthesisCamera SystemsComputational Color ConstancyCross-dataset ValidationInverse ImagingLearning-based Color ConstancyRgb Color Space
We propose a reverse camera imaging pipeline to convert arbitrary images to raw RGB responses of a specific camera. The pipeline requires only that the camera's RGB responses are characterized. The reversed pipeline helps camera developers to generate camera specific raw images and use them to train learning-based imaging pipeline algorithms. In our experiments, three recent deep color constancy architectures achieve superior results in the cross-dataset setting using generated images.
16
Noise2Noise: Learning Image Restoration without Clean Data
Jaakko Lehtinen, Jacob Munkberg, Jon Hasselgren et al. · arXiv (Cornell University) · 2018 · 988 citations · Full text
Deep joint demosaicking and denoising
Michaël Gharbi, Gaurav Chaurasia, Sylvain Paris et al. · ACM Transactions on Graphics · 2016 · 488 citations · Full text
Noise2Noise: Learning image restoration without clean data
Jaakko Lehtinen, Jacob Munkberg, Jon Hasselgren et al. · Aaltodoc (Aalto University) · 2018 · 442 citations · Full text