AI Paper Atlas

Research map / Architectures and efficiency / Neural architecture search

Single Path One-Shot Neural Architecture Search with Uniform Sampling

Zichao Guo, Xiangyu Zhang, Haoyuan Mu, Wen Heng, Zechun Liu, Yichen Wei, Jian Sun

ECCV 2020 · Architectures and efficiency › Neural architecture search · approach: One-shot and few-shot supernets · 930 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 evolutionary architecture search instead of random search on the single path supernet (ablation)SPOS · ImageNet block searchtop-1 acc (%): 73.8 -> 74.3Improved
search building blocks instead of fixed all choice_3SPOS · ImageNettop-1 acc (%): 73.4 -> 74.3Improved
add channel search on all choice_3SPOS · ImageNettop-1 acc (%): 73.4 -> 73.9Improved
add channel search on top of block searchSPOS · ImageNettop-1 acc (%): 74.3 -> 74.7Improved
search channels instead of random channel selection (ablation)SPOS · ImageNettop-1 acc (%): 73.1 -> 73.9Improved
run SPOS at same latency as FBNet-BSPOS · ImageNettop-1 acc (%): 74.1 -> 74.8Improved
run SPOS at same FLOPs as FBNet-ASPOS · ImageNettop-1 acc (%): 73.0 -> 73.2Improved
run SPOS at same latency as Proxyless (GPU)SPOS · ImageNettop-1 acc (%): 74.7 -> 75.3Improved
search mixed-precision bit widths instead of uniform 2W2ASPOS · ResNet-18 quantizationtop-1 acc (%): 65.6 -> 66.4Improved
search mixed-precision instead of uniform 3W3ASPOS · ResNet-34 quantizationtop-1 acc (%): 72.5 -> 73.9Improved
search mixed precision instead of DNAS at 4W4ASPOS · ResNet-18 quantizationtop-1 acc (%): 70.6 -> 70.5Mixed
reduce NAS-Bench-201 search space by deleting both 1x1 conv and avg pooling (Reduce-3) (ablation)SPOS supernet · NAS-Bench-201 CIFAR-10Kendall tau: 0.55 -> 0.64Improved

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