AI Paper Atlas

Research map / 3D vision / Depth and stereo

A Lightweight Approach for On-The-Fly Reflectance Estimation

Kihwan Kim, Jinwei Gu, Stephen Tyree, Pavlo Molchanov, Matthias Niessner, Jan Kautz

ICCV 2017 · 3D vision › Depth and stereo · approach: Neural light field and illumination imaging · 42 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
add the Ec regularization to Grouplet with RMSE1 loss (ablation)Grouplet · SynBRDF testRMSE: 0.432 -> 0.455Hurt
add the Ec regularization to Grouplet with RMSE1 loss (ablation)Grouplet · SynBRDF user studyuser rank: 2 -> 1Improved
add the Ec regularization to HemiCNN with CubeRoot loss (ablation)HemiCNN · SynBRDF testRMSE: 0.439 -> 0.583Hurt
add the Ec regularization to HemiCNN with CubeRoot loss (ablation)HemiCNN · SynBRDF user studyuser rank: 4 -> 6Hurt
add the Ec regularization when training on real RGBD scans (ablation)HemiCNN and Grouplet · real data (KinectFusion scans)qualitative plausibilityImproved
switch HemiCNN loss from RMSE1 to RMSE2 (Lab-based representation) (ablation)HemiCNN · SynBRDF testRMSE: 0.419 -> 0.564Hurt
switch Grouplet loss from RMSE1 to RMSE2 (ablation)Grouplet · SynBRDF user studyuser rank: 2 -> 7Hurt
choice of distance loss Ed (RMSE1, RMSE2, CubeRoot) (ablation)HemiCNN and Grouplet · SynBRDF testRMSE and gloss accuracyMixed
increase number of observations per voxel (ablation)HemiCNN, Grouplet-fast, Grouplet-slow · SynBRDF testerror: saturates around 30Improved
use more voxel samples (Grouplet-slow, 354 voxels, instead of fast, 20) (ablation)Grouplet · SynBRDFinference time: 5ms -> 90msHurt
use Grouplet instead of HemiCNNBRDF estimator · SynBRDFgloss vs diffuse albedo estimationMixed

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