AI Paper Atlas

Research map / 3D vision / Depth and stereo

Guiding Monocular Depth Estimation Using Depth-Attention Volume

Lam Huynh, Phong Nguyen-Ha, Jiri Matas, Esa Rahtu, Janne Heikkilä

ECCV 2020 · 3D vision › Depth and stereo · approach: Dense depth prediction and completion · 163 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 DAV module (ablation)depth network · NYU-Depth-v2REL: 0.140 -> 0.108Improved
add the DAV module (ablation)depth network · NYU-Depth-v2delta1: 0.827 -> 0.882Improved
increase planes from 0 to 3 (ablation)DAV · NYU-Depth-v2REL: 0.116 -> 0.110Improved
increase planes from 3 to 5 (ablation)DAV · NYU-Depth-v2REL: 0.110 -> 0.108Improved
increase planes from 5 to 7 (ablation)DAV · NYU-Depth-v2REL: 0.108 -> 0.111Hurt
increase planes from 5 to 7 (ablation)DAV · NYU-Depth-v2delta1: 0.882 -> 0.851Hurt
add the attention loss during training (ablation)DAV · NYU-Depth-v2REL: 0.126 -> 0.108Improved
keep attention loss to the end instead of dropping it after convergence (ablation)DAV · NYU-Depth-v2REL: 0.109 -> 0.108Improved
use cross-modulation in two embedding spaces instead of single embedding (ablation)DAV module · NYU-Depth-v2REL: 0.115 -> 0.108Improved
add the DAV module (cross-dataset generalization) (ablation)depth network · SUN-RGBD (trained on NYU)REL: 0.254 -> 0.238Improved
use DAV model instead of Chen'19 on ScanNetdepth network · ScanNet official testREL: 0.134 -> 0.118Improved

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