AI Paper Atlas

Research map / 3D vision / Depth and stereo

Spatio-Focal Bidirectional Disparity Estimation From a Dual-Pixel Image

Donggun Kim, Hyeonjoong Jang, Inchul Kim, Min H. Kim

CVPR 2023 · 3D vision › Depth and stereo · approach: Depth from light field and defocus · 11 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 the proposed self-supervised bidirectional disparity method instead of Wadhwa et al.Dual-pixel disparity estimation · Dataset by [23]RMSE: 4.1113 -> 3.7943Improved
use the proposed self-supervised bidirectional disparity method instead of Wadhwa et al.Dual-pixel disparity estimation · Dataset by [2]RMSE: 6.7384 -> 5.5075Improved
use the proposed method instead of the best baselineDual-pixel depth estimation · Punnappurath et al. datasetAI(1): 0.0449 -> 0.0391Improved
use the proposed method instead of the best baselineDual-pixel depth estimation · Punnappurath et al. datasetAI(2): 0.0724 -> 0.0682Improved
use the proposed method instead of the best baselineDual-pixel depth estimation · Punnappurath et al. dataset1 − |ρs|: 0.2703 -> 0.2619Improved
add the kernel-split symmetry loss (ablation)Pretrained stereo network with photometric and smoothness losses · Punnappurath et al. datasetAI(1): 0.0502 -> 0.0391Improved
add the photometric loss (ablation)Pretrained stereo network with kernel-split symmetry and smoothness losses · Punnappurath et al. datasetAI(1): 0.0689 -> 0.0391Improved
add the smoothness loss (ablation)Pretrained stereo network with kernel-split symmetry and photometric losses · Punnappurath et al. datasetAI(1): 0.0398 -> 0.0391Improved
use only the kernel-split symmetry loss (ablation)Pretrained stereo network · Punnappurath et al. datasetAI(1): 0.0826 -> 0.0660Improved
use only the photometric loss (ablation)Pretrained stereo network · Punnappurath et al. datasetAI(1): 0.0826 -> 0.0656Improved
use only the smoothness loss (ablation)Pretrained stereo network · Punnappurath et al. datasetAI(1): 0.0826 -> 0.1122Hurt

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