AI Paper Atlas

Research map / 3D vision / 3D generation

Geometry-guided Feature Learning and Fusion for Indoor Scene Reconstruction

Ruihong Yin, Sezer Karaoglu, Theo Gevers

ICCV 2023 · 3D vision › 3D generation · approach: Learning 3D reconstruction from images · 8 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 geometry integration (G2FL + G2AFF + C3NL) to NeuralReconNeuralRecon · ScanNetF-score: 0.560 -> 0.619Improved
add geometry integration to VoRTXVoRTX · ScanNetF-score: 0.598 -> 0.630Improved
add projected surface normals to the Transformer-based feature learning (ablation)NeuralRecon + G2FL · ScanNetF-score: 0.566 -> 0.587Improved
put normal and viewing priors in separate modules rather than the same Transformer (ablation)G2FL · ScanNetF-score: 0.590 -> 0.598Improved
add projected depth to G2FL (ablation)G2FL · ScanNetF-score: 0.598 -> 0.602Improved
use our geometric priors instead of SimpleRecon geometry (ablation)G2FL · ScanNetF-score: 0.571 -> 0.602Improved
add geometry-guided adaptive fusion (G2AFF) on top of G2FL (ablation)NeuralRecon + G2FL · ScanNetF-score: 0.602 -> 0.614Improved
use projective TSDF as occlusion approximation instead of projective occupancy (ablation)G2AFF · ScanNetF-score: 0.610 -> 0.609Hurt
apply 3D normal loss to all voxels without consistency weights or boundary mask (ablation)NeuralRecon + G2FL + G2AFF · ScanNetF-score: 0.614 -> 0.609Hurt
add boundary mask and consistency weight (indicator) to normal loss (ablation)C3NL · ScanNetF-score: 0.609 -> 0.619Improved
use indicator function instead of Gaussian weight in normal loss (ablation)C3NL · ScanNetF-score: 0.615 -> 0.619Improved
add NeRF-like encoding of viewing direction and depth (ablation)G2FL · ScanNetF-score: 0.600 -> 0.602Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 3 of 3 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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