Research map / Architectures and efficiency / Neural architecture search
ProxylessNAS: Direct Neural Architecture Search on Target Task and Hardware
Read the paperCheck 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 changed | Where | Measured | Effect |
|---|---|---|---|
| Add latency regularization loss during architecture search (ablation) | Proxyless-G · ImageNet, mobile latency constraint ≤ 80ms | Top-1 (%): 71.8 -> 74.2 | Helped |
| Add latency regularization loss during architecture search (ablation) | Proxyless-G · ImageNet, mobile latency constraint ≤ 80ms | Mobile latency: 83ms -> 79ms | Helped |
| Search directly on the target task rather than transfer a proxy-task architecture | Proxyless-G · CIFAR-10 test | test error (%): 2.13 -> 2.08 | Helped |
| Specialize the architecture for GPU hardware | ProxylessNAS · ImageNet | GPU latency: 6.1ms -> 5.1ms | Helped |
| Specialize the architecture for the target hardware instead of reusing a model searched for another platform | ProxylessNAS · ImageNet, CPU and mobile hardware | latency: 204.9ms vs 138.7ms; 164.1ms vs 78ms | Mixed |
Closest papers by meaning
- EcoNAS: Finding Proxies for Economical Neural Architecture Search (CVPR 2020 · similarity 0.93)
- Neural Architecture Search on ImageNet in Four GPU Hours: A Theoretically Inspired Perspective (ICLR 2021 · similarity 0.91)
- NASTransfer: Analyzing Architecture Transferability in Large Scale Neural Architecture Search (AAAI 2021 · similarity 0.91)
- Zero-Cost Proxies for Lightweight NAS (ICLR 2021 · similarity 0.90)
- DSNAS: Direct Neural Architecture Search Without Parameter Retraining (CVPR 2020 · similarity 0.90)
- Training-free Neural Architecture Search through Variance of Knowledge of Deep Network Weights (CVPR 2025 · similarity 0.90)
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.