arXiv Open Access 2022

Prior-Guided One-shot Neural Architecture Search

Peijie Dong Xin Niu Lujun Li Linzhen Xie Wenbin Zou +3 lainnya

Lihat Sumber

Abstrak

Neural architecture search methods seek optimal candidates with efficient weight-sharing supernet training. However, recent studies indicate poor ranking consistency about the performance between stand-alone architectures and shared-weight networks. In this paper, we present Prior-Guided One-shot NAS (PGONAS) to strengthen the ranking correlation of supernets. Specifically, we first explore the effect of activation functions and propose a balanced sampling strategy based on the Sandwich Rule to alleviate weight coupling in the supernet. Then, FLOPs and Zen-Score are adopted to guide the training of supernet with ranking correlation loss. Our PGONAS ranks 3rd place in the supernet Track Track of CVPR2022 Second lightweight NAS challenge. Code is available in https://github.com/pprp/CVPR2022-NAS?competition-Track1-3th-solution.

Topik & Kata Kunci

cs.CV

Penulis (8)

Peijie Dong

Xin Niu

Lujun Li

Linzhen Xie

Wenbin Zou

Tian Ye

Zimian Wei

Hengyue Pan

Format Sitasi

APA MLA BibTeX

Dong, P., Niu, X., Li, L., Xie, L., Zou, W., Ye, T. et al. (2022). Prior-Guided One-shot Neural Architecture Search. https://arxiv.org/abs/2206.13329

Akses Cepat

Lihat di Sumber

Informasi Jurnal

Tahun Terbit: 2022
Bahasa: en
Sumber Database: arXiv
Akses: Open Access ✓