Fast-N-Squeeze: Towards Real-Time Spectral Reconstruction from RGB Images
Abstract
We present an efficient method for the reconstruction of multispectral information from RGB images, as part of the NTIRE 2022 Spectral Reconstruction Challenge. Given an input image, our method determines a global RGB-to-spectral linear transformation matrix, based on a search through optimal matrices from training images that share low-level features with the input. The resulting spectral signatures are then adjusted by a global scaling factor, determined through a lightweight SqueezeNet-inspired neural network. By combining the efficiency of linear transformation matrices with the data-driven effectiveness of convolutional neural networks, we are able to achieve superior performance than winners of the previous editions of the challenge.
Cite
Text
Agarla et al. "Fast-N-Squeeze: Towards Real-Time Spectral Reconstruction from RGB Images." IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, 2022. doi:10.1109/CVPRW56347.2022.00122Markdown
[Agarla et al. "Fast-N-Squeeze: Towards Real-Time Spectral Reconstruction from RGB Images." IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, 2022.](https://mlanthology.org/cvprw/2022/agarla2022cvprw-fastnsqueeze/) doi:10.1109/CVPRW56347.2022.00122BibTeX
@inproceedings{agarla2022cvprw-fastnsqueeze,
title = {{Fast-N-Squeeze: Towards Real-Time Spectral Reconstruction from RGB Images}},
author = {Agarla, Mirko and Bianco, Simone and Buzzelli, Marco and Celona, Luigi and Schettini, Raimondo},
booktitle = {IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops},
year = {2022},
pages = {1131-1138},
doi = {10.1109/CVPRW56347.2022.00122},
url = {https://mlanthology.org/cvprw/2022/agarla2022cvprw-fastnsqueeze/}
}