Unchain the Search Space with Hierarchical Differentiable Architecture Search


  • Guanting Liu Malong LLC
  • Yujie Zhong Malong LLC
  • Sheng Guo Malong LLC
  • Matthew R. Scott Malong LLC
  • Weilin Huang Malong LLC




Transfer/Adaptation/Multi-task/Meta/Automated Learning, Object Detection & Categorization


Differentiable architecture search (DAS) has made great progress in searching for high-performance architectures with reduced computational cost. However, DAS-based methods mainly focus on searching for a repeatable cell structure, which is then stacked sequentially in multiple stages to form the networks. This configuration significantly reduces the search space, and ignores the importance of connections between the cells. To overcome this limitation, in this paper, we propose a Hierarchical Differentiable Architecture Search (H-DAS) that performs architecture search both at the cell level and at the stage level. Specifically, the cell-level search space is relaxed so that the networks can learn stage-specific cell structures. For the stage-level search, we systematically study the architectures of stages, including the number of cells in each stage and the connections between the cells. Based on insightful observations, we design several search rules and losses, and mange to search for better stage-level architectures. Such hierarchical search space greatly improves the performance of the networks without introducing expensive search cost. Extensive experiments on CIFAR10 and ImageNet demonstrate the effectiveness of the proposed H-DAS. Moreover, the searched stage-level architectures can be combined with the cell structures searched by existing DAS methods to further boost the performance. Code is available at: https://github.com/msight-tech/research-HDAS




How to Cite

Liu, G., Zhong, Y., Guo, S., Scott, M. R., & Huang, W. (2021). Unchain the Search Space with Hierarchical Differentiable Architecture Search. Proceedings of the AAAI Conference on Artificial Intelligence, 35(10), 8644-8652. https://doi.org/10.1609/aaai.v35i10.17048



AAAI Technical Track on Machine Learning III