NeuralScale: Efficient Scaling of Neurons for Resource-Constrained Deep Neural Networks

TL;DR

NeuralScale introduces a method to efficiently scale neural network architectures by optimizing neuron configurations for resource constraints, leading to significant accuracy improvements across multiple models and datasets.

Contribution

The paper proposes NeuralScale, a novel approach combining parameterized neuron scaling and architecture descent to optimize neural network size for better accuracy under resource limits.

Findings

Achieved up to 8.56% accuracy improvement on MobileNetV2.

Demonstrated effectiveness across VGG11, ResNet18, and MobileNetV2.

Validated on CIFAR10, CIFAR100, and TinyImageNet datasets.

Abstract

Deciding the amount of neurons during the design of a deep neural network to maximize performance is not intuitive. In this work, we attempt to search for the neuron (filter) configuration of a fixed network architecture that maximizes accuracy. Using iterative pruning methods as a proxy, we parameterize the change of the neuron (filter) number of each layer with respect to the change in parameters, allowing us to efficiently scale an architecture across arbitrary sizes. We also introduce architecture descent which iteratively refines the parameterized function used for model scaling. The combination of both proposed methods is coined as NeuralScale. To prove the efficiency of NeuralScale in terms of parameters, we show empirical simulations on VGG11, MobileNetV2 and ResNet18 using CIFAR10, CIFAR100 and TinyImageNet as benchmark datasets. Our results show an increase in accuracy of 3.04%, 8.56% and 3.41% for VGG11, MobileNetV2 and ResNet18 on CIFAR10, CIFAR100 and TinyImageNet respectively under a parameter-constrained setting (output neurons (filters) of default configuration with scaling factor of 0.25).