AI Paper Atlas

Research map / 3D vision / Depth and stereo

Depth Completion Using Plane-Residual Representation

Byeong-Uk Lee, Kyunghyun Lee, In So Kweon

CVPR 2021 · 3D vision › Depth and stereo · approach: Dense depth prediction and completion · 58 citations (OpenAlex)

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 Plane-Residual classification plus residual regression rather than direct depth regressionOurs · NYU Depth V2RMSE: 0.118 -> 0.104Improved
use uniformly and relatively set planes (UR) rather than uniformly and absolutely set planes (UA) (ablation)Plane-Residual depth completion · NYU Depth V2RMSE: 0.113 -> 0.104Improved
use UR rather than disparity-wise and relatively set planes (DR) (ablation)Plane-Residual depth completion · NYU Depth V2RMSE: 0.114 -> 0.104Improved
use UR rather than disparity-wise and absolutely set planes (DA) (ablation)Plane-Residual depth completion · NYU Depth V2RMSE: 0.117 -> 0.104Improved
use 8 planes rather than 4 (ablation)Plane-Residual depth completion · NYU Depth V2RMSE: 0.111 -> 0.104Improved
use 8 planes rather than 16 (ablation)Plane-Residual depth completion · NYU Depth V2RMSE: 0.105 -> 0.104No clear effect
use 8 planes rather than 32 (ablation)Plane-Residual depth completion · NYU Depth V2RMSE: 0.106 -> 0.104No clear effect
add channel-wise guided image filtering of classification logits (ablation)Plane-Residual depth completion · NYU Depth V2RMSE: 0.105 -> 0.104Improved
add confidence-guided residual learning (ablation)Plane-Residual depth completion · NYU Depth V2RMSE: 0.122 -> 0.104Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 0 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