A Neural Architecture Search Method using Auxiliary Evaluation Metric based on ResNet Architecture
This work addresses neural architecture design for computer vision researchers, but it is incremental as it builds on existing ResNet frameworks and search methods.
The paper tackled neural architecture search by proposing a ResNet-based search space with optimization using both recognition accuracy and validation loss as objectives, resulting in competitive network architectures on MNIST, Fashion-MNIST, and CIFAR100 datasets.
This paper proposes a neural architecture search space using ResNet as a framework, with search objectives including parameters for convolution, pooling, fully connected layers, and connectivity of the residual network. In addition to recognition accuracy, this paper uses the loss value on the validation set as a secondary objective for optimization. The experimental results demonstrate that the search space of this paper together with the optimisation approach can find competitive network architectures on the MNIST, Fashion-MNIST and CIFAR100 datasets.