AI Paper Atlas

Research map / Image generation / Image restoration

Computational Flash Photography Through Intrinsics

Sepideh Sarajian Maralan, Chris Careaga, Yagiz Aksoy

CVPR 2023 · Image generation › Image restoration · approach: Reflection removal · 3 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
use intrinsic albedo/shading modeling instead of direct relit-image regression (ablation)Flash decomposition · Flash decomposition test setSSIM: 0.774 -> 0.8553Improved
use intrinsic albedo/shading modeling instead of direct relit-image regression (ablation)Flash generation · Flash generation test setPSNR: 14.5329 -> 19.1863Improved
add cycle loss (ablation)Flash generation · Flash generation test setSSIM: 0.8597 -> 0.8601No clear effect
add cycle loss (ablation)Flash generation · Flash generation test setIS: 6.4777 -> 7.3730Improved
add multi-scale gradient loss (ablation)Flash decomposition · Flash decomposition test setPSNR: 16.2628 -> 19.6044Improved
add multi-scale gradient loss (ablation)Flash generation · Flash generation test setIS: 6.5127 -> 7.3730Improved
add depth and surface-normal inputs (ablation)Flash decomposition · Flash decomposition test setPSNR: 14.5056 -> 19.6044Improved
add depth and surface-normal inputs (ablation)Flash generation · Flash generation test setIS: 6.5307 -> 7.3730Improved

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