AI Paper Atlas

Research map / 3D vision / Depth and stereo

SeeGroup: Multi-Layer Depth Estimation of Transparent Surfaces via Self-Determined Grouping

Hongyu Wen, Jia Deng

CVPR 2026 · 3D vision › Depth and stereo · approach: Depth from light field and defocus · 0 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 SeeGroup (RD + max-mixture intensity + permutation-invariant loss) over Multi-head headDepth Anything V2 multi-layer depth · LayeredDepth test (All)quadruplet accuracy: 61.34 -> 70.09Improved
add recurrent decomposition instead of GRU (ablation)SeeGroup · LayeredDepth val (ViT-L)quadruplet accuracy All: 67.82 -> 71.50Improved
use recurrent decomposition instead of independent multi-head on deep layers (ablation)SeeGroup · LayeredDepth val Layer 5quadruplet accuracy: 39.64 -> 66.14Improved
use recurrent decomposition instead of multi-head overall (ablation)SeeGroup · LayeredDepth val (All)quadruplet accuracy: 73.04 -> 71.50Mixed
use max-mixture instead of weighted mixture (ablation)SeeGroup (ViT-S) · LayeredDepth val (All)quadruplet accuracy: 56.12 -> 69.03Improved
use max-mixture (permutation invariant) instead of depth-ordered sorted mixture (ablation)SeeGroup (ViT-S) · LayeredDepth val (All)quadruplet accuracy: 52.68 -> 69.03Improved
add coverage loss to intensity and GM losses (ablation)SeeGroup (ViT-S) · LayeredDepth val (All)quadruplet accuracy: 68.36 -> 69.03Improved
add gradient matching loss to L1 (ablation)L1 loss · LayeredDepth val (All)quadruplet accuracy: 42.49 -> 57.00Improved
use intensity loss instead of L1 with GM (ablation)SeeGroup (ViT-S) · LayeredDepth val (All)quadruplet accuracy: 57.00 -> 68.36Improved
use full loss instead of SiLog (ablation)SeeGroup (ViT-S) · LayeredDepth val (All)quadruplet accuracy: 48.20 -> 69.03Improved
use SeeGroup instead of Multi-head DA v2 on synthetic in-domain dataSeeGroup · LayeredDepth-Syn val Layer 1AbsRel: 8.24 -> 9.82Hurt
use RD instead of GRU for model size (ablation)RD · model size#Param: 731M -> 356MImproved

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