AI Paper Atlas

Research map / 3D vision / Depth and stereo

Hyper-Skin: A Hyperspectral Dataset for Reconstructing Facial Skin-Spectra from RGB Images

Pai Chet Ng, Zhixiang Chi, Yannick Verdie, Juwei Lu, Konstantinos N Plataniotis

NeurIPS 2023 · 3D vision › Depth and stereo · approach: Snapshot hyperspectral imaging · 4 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
re-train on Hyper-Skin instead of NTIRE2022 pre-trained weightsHSCNN · Hyper-Skin (RGB, VIS) with backgroundSSIM: 0.683 ± 0.027 -> 0.916 ± 0.013Improved
re-train on Hyper-Skin instead of NTIRE2022 pre-trained weightsMST++ · Hyper-Skin (RGB, VIS) with backgroundSSIM: 0.602 ± 0.042 -> 0.923 ± 0.011Improved
re-train on Hyper-Skin instead of NTIRE2022 pre-trained weightsHRNet · Hyper-Skin (RGB, VIS) with backgroundSAM: 0.648 ± 0.062 -> 0.147 ± 0.014Improved
re-train on Hyper-Skin instead of NTIRE2022 pre-trained weightsMST++ · Hyper-Skin (RGB, VIS) w/o backgroundSAM: 0.628 ± 0.050 -> 0.107 ± 0.010Improved
remove background with a mask when evaluating (ablation)MST++ pre-trained · Hyper-Skin (RGB, VIS)SSIM: 0.602 ± 0.042 -> 0.766 ± 0.035Improved
remove background with a mask when evaluating (ablation)HSCNN re-trained · Hyper-Skin (RGB, VIS)SSIM: 0.916 ± 0.013 -> 0.950 ± 0.011Improved
remove background with a mask when evaluating (ablation)HRNet re-trained · Hyper-Skin (MSI, NIR)SAM: 0.094 ± 0.009 -> 0.086 ± 0.010Improved
use (MSI, NIR) 4-channel input instead of (RGB, VIS)MST++ re-trained · Hyper-Skin with backgroundSAM: 0.113 ± 0.009 -> 0.086 ± 0.006Improved
use (MSI, NIR) 4-channel input instead of (RGB, VIS)HRNet re-trained · Hyper-Skin with backgroundSSIM: 0.933 ± 0.021 -> 0.955 ± 0.006Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy