AI Paper Atlas

Research map / 3D vision / Depth and stereo

Monocular Scene Reconstruction with 3D SDF Transformers

Weihao Yuan, Xiaodong Gu, Heng Li, Zilong Dong, Siyu Zhu

ICLR 2023 · 3D vision › Depth and stereo · approach: Feed-forward 3D reconstruction · 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
Replace the 3D CNN U-Net baseline with the SDF transformer (ablation)ScanNet reconstruction with variance fusion · ScanNet ablationF-score: 0.647 -> 0.732Improved
Add variance-based feature fusion (ablation)3D CNN U-Net reconstruction baseline · ScanNet ablationF-score: 0.636 -> 0.647Improved
Add bottom-level global attention and global context encoding (ablation)SDF transformer · ScanNet ablationF-score: 0.732 -> 0.741Improved
Add post-dilate-attention after the down-up flow (ablation)SDF transformer with global module · ScanNet ablationF-score: 0.741 -> 0.754Improved
Increase sparse attention window size from 1 to 10 (ablation)SDF transformer · ScanNet ablationF-score: 0.656 -> 0.754Improved
Increase sparse attention window size from 8 to 10 (ablation)SDF transformer · ScanNet ablationF-score: 0.746 -> 0.754Improved
Add translation pose noise with σT 1 cm to all frames (ablation)Ours · ScanNet robustnessF-score: 0.705 -> 0.663Hurt
Add rotation pose noise with σR 1 deg to all frames (ablation)Ours · ScanNet robustnessF-score: 0.705 -> 0.351Hurt
Reduce voxel size from 4cm to 2cm (ablation)Ours on Scene-709 · Scene-709F-score: 0.780 -> 0.832Improved
Use the proposed 3D SDF transformer instead of the previous best methodScanNet mesh reconstruction with occlusion-mask evaluation · ScanNet testF-score: 0.655 -> 0.754Improved
Use the proposed method instead of VoRTXScanNet rendered depth · ScanNet testAbs Rel: 0.061 -> 0.051Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 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