AI Paper Atlas

Research map / 3D vision / Depth and stereo

FlashDepth: Real-time Streaming Video Depth Estimation at 2K Resolution

Gene Chou, Wenqi Xian, Guandao Yang, Mohamed Abdelfattah, Bharath Hariharan, Noah Snavely, Ning Yu, Paul Debevec

ICCV 2025 · 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
add UNet super-resolution on high-res input (ablation)FlashDepth-L · Unreal4Kdelta1: 0.566 -> 0.490Hurt
add UNet super-resolution on high-res input (ablation)FlashDepth-L · Unreal4KF1: 0.048 -> 0.034Hurt
add DAGF depth super-resolution (ablation)FlashDepth-L · Unreal4Kdelta1: 0.566 -> 0.377Hurt
add DAGF depth super-resolution (ablation)FlashDepth-L · Unreal4KFPS: 30 -> 7.7Hurt
use hybrid cross-attention model instead of bilinear upsampling (ablation)FlashDepth-L · Unreal4KF1: 0.048 -> 0.109Improved
use hybrid cross-attention model instead of bilinear upsampling (ablation)FlashDepth-L · UrbanSynF1: 0.136 -> 0.185Improved
process frames natively at 2K instead of 518pxFlashDepth-L · Unreal4KF1: 0.048 -> 0.143Improved
process frames natively at 2K instead of 518pxFlashDepth-L · Unreal4KFPS: 30 -> 6.0Hurt
use hybrid with FlashDepth-L features instead of FlashDepth-S alone (ablation)FlashDepth-S · UrbanSyndelta1: 0.784 -> 0.862Improved
use large model instead of small model (ablation)FlashDepth · ETH3Ddelta1: 0.786 -> 0.875Improved
replace Mamba with Transformer recurrent module (ablation)FlashDepth-L · Unreal4Kdelta1: 0.566 -> 0.608Improved
replace Mamba with Transformer recurrent module (ablation)FlashDepth-L · Sinteldelta1: 0.642 -> 0.615Hurt

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