AI Paper Atlas

Research map / Point clouds / Segmentation and detection

Tangent Convolutions for Dense Prediction in 3D

Maxim Tatarchenko, Jaesik Park, Vladlen Koltun, Qian-Yi Zhou

CVPR 2018 · Point clouds › Segmentation and detection · approach: Scene completion and instance segmentation · 618 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 tangent convolutions rather than OctNetSemantic scene segmentation · ScanNetmIoU: 18.1 -> 40.9Improved
Use tangent convolutions rather than PointNetSemantic scene segmentation · S3DISmIoU: 41.3 -> 52.8Improved
Use tangent convolutions rather than SnapNetSemantic scene segmentation · Semantic3DmIoU: 67.7 -> 66.4Hurt
Add RGB input to distance, height, and normals (ablation)Tangent convolution network · Semantic3DmIoU: 52.5 -> 66.4Improved
Add RGB input to distance, height, and normals (ablation)Tangent convolution network · S3DISmIoU: 51.7 -> 52.8Improved
Add RGB input to distance, height, and normals (ablation)Tangent convolution network · ScanNetmIoU: 40.7 -> 40.9No clear effect
Add height input to distance (ablation)Tangent convolution network · Semantic3DmIoU: 58.1 -> 58.0No clear effect
Add height input to distance (ablation)Tangent convolution network · ScanNetmIoU: 40.9 -> 40.3Hurt
Add height input to distance (ablation)Tangent convolution network · S3DISmIoU: 49.8 -> 50.0No clear effect
Add normals to distance and height (ablation)Tangent convolution network · Semantic3DmIoU: 58.0 -> 52.5Hurt
Add normals to distance and height (ablation)Tangent convolution network · ScanNetmIoU: 40.3 -> 40.7No clear effect
Add normals to distance and height (ablation)Tangent convolution network · S3DISmIoU: 50.0 -> 51.7Improved

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