Fine-tuning differentiable architecture search for image classification

Abstract

Neural Architecture Search (NAS) has gained attraction due to superior classification performance. Differentiable Architecture Search (DARTS) is a popular NAS method because of its computational efficiency. To limit computational resources DARTS makes numerous approximations. These approximations result in inferior performance. We propose to fine-tune DARTS using fixed operations as they are independent of these approximations. Our method offers a good trade-off between the number of parameters and classification accuracy. Furthermore, based on the influence of input on the performance of a neural network, we propose a dual-stem approach instead of a single-stem approach. Our approach improves the top-1 accuracy on Fashion-MNIST, CompCars and MIO-TCD datasets by 0.56%, 0.50%, and 0.39%, respectively compared to the state-of-the-art approaches. Our approach performs better than DARTS, improving the accuracy by 0.28%, 1.64%, 0.34%, 4.5%, and 3.27% compared to DARTS, on CIFAR-10, CIFAR-100, Fashion-MNIST, Com-pCars, and MIO-TCD datasets, respectively.