AI Paper Atlas

Research map / Image generation / Autoregressive and latent generation

Scaling Dense Prediction with Latent Decoding

Xiaoyang Wu, Yixing Lao, Chengyao Wang, Senqiao Yang, Yujia Zhang, Hengshuang ZHAO

ECCV 2026 · Image generation › Autoregressive and latent generation · approach: High-resolution diffusion synthesis · 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
replace fixed-schedule fusion with stage-wise adaptive selection of encoder features (ablation)L-DPT · Hypersim val relative depthARel: 9.67 -> 8.89Improved
use stage-specific selectors instead of one shared selector (ablation)L-DPT · Hypersim val relative depthdelta1: 92.06 -> 92.76Improved
replace DPT decoder with L-DPT S (23M vs 230M) on frozen DINOv3 ViT-7BDINOv3 depther · DIODE zero-shotARel: 25.6 -> 22.6Improved
replace DPT decoder with L-DPT S on frozen DINOv3 ViT-7BDINOv3 depther · NYUv2 zero-shotARel: 4.3 -> 5.0Hurt
scale L-DPT decoder from S to L (23M to 269M) (ablation)L-DPT · DIODE zero-shotARel: 22.6 -> 22.0Improved
replace DINOv3 ViT-Adapter + Mask2Former with L-DPT S decoderDINOv3 ViT-7B segmentor · PASCAL VOC 2012 (simple)mIoU: 90.1 -> 93.4Improved
replace DINOv3 segmentor with L-DPT L on frozen ViT-7BDINOv3 ViT-7B segmentor · COCO-Stuff 164k (simple)mIoU: 53.8 -> 54.0Improved
use latent decoding instead of DPT decoder at high resolutiondense decoder · decoder peak memory at high end of resolution sweeppeak memory (GB): 59.4 -> 15.3Improved
use latent decoding instead of DPT decoder at high resolutiondense decoder · decoder latency at high end of resolution sweeplatency (s): 394 -> 5.3Improved
use L-DPT-L instead of InfiniDepth with DINOv3-Lrelative depth · ETH3D delta1delta1: 98.6 -> 97.4Hurt

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