Rethinking the Number of Channels for the Convolutional Neural Network
This work addresses the bottleneck of channel width optimization in neural architecture search for researchers and practitioners, offering a domain-specific incremental improvement.
The paper tackles the problem of efficiently designing the number of channels in convolutional neural networks through an automatic architecture search method, achieving accuracy improvements of about 0.5% on CIFAR-10 and 2.33% on CIFAR-100 with fewer parameters and minimal computational resources (0.4-1.3 GPU-days).
Latest algorithms for automatic neural architecture search perform remarkable but few of them can effectively design the number of channels for convolutional neural networks and consume less computational efforts. In this paper, we propose a method for efficient automatic architecture search which is special to the widths of networks instead of the connections of neural architecture. Our method, functionally incremental search based on function-preserving, will explore the number of channels rapidly while controlling the number of parameters of the target network. On CIFAR-10 and CIFAR-100 classification, our method using minimal computational resources (0.4~1.3 GPU-days) can discover more efficient rules of the widths of networks to improve the accuracy by about 0.5% on CIFAR-10 and a~2.33% on CIFAR-100 with fewer number of parameters. In particular, our method is suitable for exploring the number of channels of almost any convolutional neural network rapidly.