Deep Hyperspectral Prior: Denoising, Inpainting, Super-Resolution

1 Feb 2019  ·  Oleksii Sidorov, Jon Yngve Hardeberg ·

Deep learning algorithms have demonstrated state-of-the-art performance in various tasks of image restoration. This was made possible through the ability of CNNs to learn from large exemplar sets. However, the latter becomes an issue for hyperspectral image processing where datasets commonly consist of just a few images. In this work, we propose a new approach to denoising, inpainting, and super-resolution of hyperspectral image data using intrinsic properties of a CNN without any training. The performance of the given algorithm is shown to be comparable to the performance of trained networks, while its application is not restricted by the availability of training data. This work is an extension of original "deep prior" algorithm to HSI domain and 3D-convolutional networks.

PDF Abstract


Task Dataset Model Metric Name Metric Value Global Rank Result Benchmark
Hyperspectral Image Denoising HYDICE DC Mall Deep HS (prior 3D) MPSNR 23.24 # 1
MSSIM 0.852 # 1
SAM 9.91 # 1
Hyperspectral Image Inpainting Indian Pines Deep HS (prior 3D) MPSNR 35.34 # 1
MSSIM 0.966 # 1
SAM 1.133 # 1
Hyperspectral Image Super-Resolution ROSIS-03 Deep HS (prior 3D) MPSNR 32.31 # 1
MSSIM 0.945 # 1
SAM 4.692 # 1


No methods listed for this paper. Add relevant methods here