AI Paper Atlas

Research map / 3D vision / Depth and stereo

LiteSense: Lifting Lightweight ToF with RGB for High-Resolution Metric Depth Estimation

Yusheng Li, Lizhi Lou, Yan Tang, Zekai Miao, Shaoming Zhang, Jianmei Wang

CVPR 2026 · 3D vision › Depth and stereo · approach: Dense depth prediction and completion

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 ToF depth and CNH histogram with PCSI fusion to an RGB-only network (ablation)lightweight RGB baseline · NYUv2delta1 / AbsRel: 0.663 -> 0.982; 0.216 -> 0.029Improved
use native CNH distribution instead of depth-only or DELTAR sampled distribution (ablation)LiteSense auxiliary input · NYUv2accuracyImproved
concatenate upsampled ToF depth to RGB as an input channel (ablation)LiteSense · NYUv2accuracyImproved
replace patch-wise PCSI with global cross-attention between RGB-D and CNH features (ablation)LiteSense · NYUv2accuracyHurt
fuse CNH only at 1/2 and 1/4 decoder scales rather than all stagesLiteSense decoder · NYUv2artifactsImproved
use LiteSense instead of DELTAR for RGB-ToF fusionDELTAR · NYUv2delta1 / AbsRel: delta1 0.952 -> 0.982; AbsRel 0.064 -> 0.029Improved
use LiteSense instead of DELTARDELTAR · NYUv2parameters: 18.55 -> 5.48 MImproved
add ToF auxiliary priors to metric depth estimation (zero-shot)monocular baselines · SUN RGB-D zero-shotdelta1: Metric3D-V2-L 0.962 -> LiteSense 0.980Improved
fine-tune on real sensor data instead of simulated inputsLiteSense · THDR3Kdelta1 / AbsRel: delta1 0.959 -> 0.914; AbsRel 0.054 -> 0.091Hurt

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