DeepAI AI Chat
Log In Sign Up

Towards Accurate and Compact Architectures via Neural Architecture Transformer

by   Yong Guo, et al.
South China University of Technology International Student Union

Designing effective architectures is one of the key factors behind the success of deep neural networks. Existing deep architectures are either manually designed or automatically searched by some Neural Architecture Search (NAS) methods. However, even a well-designed/searched architecture may still contain many nonsignificant or redundant modules/operations. Thus, it is necessary to optimize the operations inside an architecture to improve the performance without introducing extra computational cost. To this end, we have proposed a Neural Architecture Transformer (NAT) method which casts the optimization problem into a Markov Decision Process (MDP) and seeks to replace the redundant operations with more efficient operations, such as skip or null connection. Note that NAT only considers a small number of possible transitions and thus comes with a limited search/transition space. As a result, such a small search space may hamper the performance of architecture optimization. To address this issue, we propose a Neural Architecture Transformer++ (NAT++) method which further enlarges the set of candidate transitions to improve the performance of architecture optimization. Specifically, we present a two-level transition rule to obtain valid transitions, i.e., allowing operations to have more efficient types (e.g., convolution->separable convolution) or smaller kernel sizes (e.g., 5x5->3x3). Note that different operations may have different valid transitions. We further propose a Binary-Masked Softmax (BMSoftmax) layer to omit the possible invalid transitions. Extensive experiments on several benchmark datasets show that the transformed architecture significantly outperforms both its original counterpart and the architectures optimized by existing methods.


page 1

page 2

page 3

page 4


NAT: Neural Architecture Transformer for Accurate and Compact Architectures

Designing effective architectures is one of the key factors behind the s...

Neural Architecture Search on Efficient Transformers and Beyond

Recently, numerous efficient Transformers have been proposed to reduce t...

Efficient Neural Architecture Search with Network Morphism

While neural architecture search (NAS) has drawn increasing attention fo...

Contrastive Neural Architecture Search with Neural Architecture Comparators

One of the key steps in Neural Architecture Search (NAS) is to estimate ...

ENAS U-Net: Evolutionary Neural Architecture Search for Retinal Vessel Segmentation

The accurate retina vessel segmentation (RVS) is of great significance t...

Neural Architecture Search for Speech Emotion Recognition

Deep neural networks have brought significant advancements to speech emo...

Theory-Inspired Path-Regularized Differential Network Architecture Search

Despite its high search efficiency, differential architecture search (DA...