Neural Architecture Search (NAS) is a powerful tool to automatically design deep neural networks for many tasks, including image classification. Due to the significant computational burden of the search phase, most NAS methods have focused so far on small, balanced datasets. All attempts at conducting NAS at large scale have employed small proxy sets, and then transferred the learned architectures to larger datasets by replicating or stacking the searched cells. We propose a NAS method based on polyharmonic splines that can perform search directly on large scale, imbalanced target datasets. We demonstrate the effectiveness of our method on the ImageNet22K benchmark[16], which contains 14 million images distributed in a highly imbalanced manner over 21,841 categories. By exploring the search space of the ResNet [23] and Big-Little Net ResNext [11] architectures directly on ImageNet22K, our polyharmonic splines NAS method designed a model which achieved a top-1 accuracy of 40.03% on ImageNet22K, an absolute improvement of 3.13% over the state of the art with similar global batch size [15].
@article{arxiv.2011.10608,
title = {Large Scale Neural Architecture Search with Polyharmonic Splines},
author = {Ulrich Finkler and Michele Merler and Rameswar Panda and Mayoore S. Jaiswal and Hui Wu and Kandan Ramakrishnan and Chun-Fu Chen and Minsik Cho and David Kung and Rogerio Feris and Bishwaranjan Bhattacharjee},
journal= {arXiv preprint arXiv:2011.10608},
year = {2020}
}