학술논문
NTIRE 2018 Challenge on Spectral Reconstruction from RGB Images
Document Type
Conference
Author
Arad, B.; Ben-Shahar, O.; Timofte, R.; Van Gool, L.; Zhang, L.; Yang, M.-H.; Xiong, Z.; Chen, C.; Shi, Z.; Liu, D.; Wu, F.; Lanaras, C.; Galliani, S.; Schindler, K.; Stiebel, T.; Koppers, S.; Seltsam, P.; Zhou, R.; El Helou, M.; Lahoud, F.; Shahpaski, M.; Zheng, K.; Gao, L.; Zhang, B.; Cui, X.; Yu, H.; Can, Y.B.; Alvarez-Gila, A.; Van de Weijer, J.; Garrote, E.; Galdran, A.; Sharma, M.; Koundinya, S.; Upadhyay, A.; Manekar, R.; Mukhopadhyay, R.; Sharma, H.; Chaudhury, S.; Nagasubramanian, K.; Ghosal, S.; Singh, A.K.; Singh, A.; Ganapathysubramanian, B.; Sarkar, S.
Source
2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) CVPRW Computer Vision and Pattern Recognition Workshops (CVPRW), 2018 IEEE/CVF Conference on. :1042-104209 Jun, 2018
Subject
Language
ISSN
2160-7516
Abstract
This paper reviews the first challenge on spectral image reconstruction from RGB images, i.e., the recovery of whole-scene hyperspectral (HS) information from a 3-channel RGB image. The challenge was divided into 2 tracks: the "Clean" track sought HS recovery from noiseless RGB images obtained from a known response function (representing spectrally-calibrated camera) while the "Real World" track challenged participants to recover HS cubes from JPEG-compressed RGB images generated by an unknown response function. To facilitate the challenge, the BGU Hyperspectral Image Database [4] was extended to provide participants with 256 natural HS training images, and 5+10 additional images for validation and testing, respectively. The "Clean" and "Real World" tracks had 73 and 63 registered participants respectively, with 12 teams competing in the final testing phase. Proposed methods and their corresponding results are reported in this review.