atlas

Research map / Detection and segmentation / CNN features

Auto-DeepLab: Hierarchical Neural Architecture Search for Semantic Image Segmentation

CVPR 2019 · Detection and segmentation › CNN features · approach: Convolution-transformer backbones · 1,058 citations (OpenAlex)

Check your idea against 108,967 papers

AI-extracted summary. Written by AI from the paper's title only (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
Increase filter multiplier from 20 to 48Auto-DeepLab · Cityscapes validationmIOU (%): 79.74 -> 80.33Helped
Increase training iterations from 500K to 1.5M (ablation)Auto-DeepLab-S · Cityscapes validationmIOU (%): 75.20 -> 78.00; 2.8%Helped
Add Scheduled Drop Path (ablation)Auto-DeepLab-S · Cityscapes validationmIOU (%): 78.00 -> 79.74; 1.74%Helped
Use coarse annotationsAuto-DeepLab-L · Cityscapes testmIOU (%): 80.4 -> 82.1Helped
Use multi-scale inference (ablation)Auto-DeepLab-L · PASCAL VOC 2012 validationmIOU (%): 73.76 -> 75.26Helped
Pretrain on COCO (ablation)Auto-DeepLab-L · PASCAL VOC 2012 validationmIOU (%): 73.76 -> 80.75Helped

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