AI Paper Atlas

Research map / Image generation / Super-resolution

Fight Ill-Posedness With Ill-Posedness: Single-Shot Variational Depth Super-Resolution From Shading

Bjoern Haefner, Yvain Quéau, Thomas Möllenhoff, Daniel Cremers

CVPR 2018 · Image generation › Super-resolution · approach: Perceptual and real-world SR · 44 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
Jointly use high-resolution shading and low-resolution depth rather than blind depth super-resolutionDepth super-resolution on Joyful Yell · Joyful Yell (voronoi)mean angular error on normals: 18.9° vs 42.8°Improved
Jointly use high-resolution shading and low-resolution depth rather than blind depth super-resolutionDepth super-resolution on Joyful Yell · Joyful Yell (constant)mean angular error on normals: 16.0° vs 42.1°Improved
Jointly use high-resolution shading and low-resolution depth rather than blind depth super-resolutionDepth super-resolution on Joyful Yell · Joyful Yell (rectcircle)mean angular error on normals: 20.0° vs 42.5°Improved
Jointly use high-resolution shading and low-resolution depth rather than image-guided depth upsamplingDepth super-resolution on Joyful Yell · Joyful Yell (voronoi)mean angular error on normals: 18.9° vs 34.2°Improved
Jointly use high-resolution shading and low-resolution depth rather than image-guided depth upsamplingDepth super-resolution on Joyful Yell · Joyful Yell (constant)mean angular error on normals: 16.0° vs 31.5°Improved
Jointly use high-resolution shading and low-resolution depth rather than image-guided depth upsamplingDepth super-resolution on Joyful Yell · Joyful Yell (rectcircle)mean angular error on normals: 20.0° vs 31.9°Improved
Jointly use high-resolution shading and low-resolution depth rather than shading-based refinement with downsampled colorDepth super-resolution on Joyful Yell · Joyful Yell (voronoi)mean angular error on normals: 18.9° vs 39.2°Improved
Jointly use high-resolution shading and low-resolution depth rather than shading-based refinement with downsampled colorDepth super-resolution on Joyful Yell · Joyful Yell (constant)mean angular error on normals: 16.0° vs 40.1°Improved
Jointly use high-resolution shading and low-resolution depth rather than shading-based refinement with downsampled colorDepth super-resolution on Joyful Yell · Joyful Yell (rectcircle)mean angular error on normals: 20.0° vs 48.0°Improved
Remove the shape-from-shading term (ablation)Joint depth super-resolution and shape-from-shading · Synthetic depth super-resolutionrecovery of high-frequency geometric detailHurt
Remove low-resolution depth guidance by setting µ = 0 (ablation)Joint depth super-resolution and shape-from-shading · Synthetic depth super-resolutionlow-frequency shape accuracyHurt
Increase the depth-data weight µ too far (ablation)Joint depth super-resolution and shape-from-shading · Joyful Yell synthetic parameter studydepth RMSEHurt

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