Learning Time/Memory-Efficient Deep Architectures with Budgeted Super Networks
This addresses the need for efficient neural networks in resource-constrained environments, offering a novel method for budget-aware architecture search.
The paper tackles the problem of discovering neural network architectures that are efficient in both prediction quality and computational or memory cost, proposing Budgeted Super Networks (BSN) that learn under a budgeted objective. It shows that BSN can discover architectures with better accuracy than ResNet and Convolutional Neural Fabrics on CIFAR-10 and CIFAR-100 at lower costs, such as less than 100 milliseconds or 50 Mb memory.
We propose to focus on the problem of discovering neural network architectures efficient in terms of both prediction quality and cost. For instance, our approach is able to solve the following tasks: learn a neural network able to predict well in less than 100 milliseconds or learn an efficient model that fits in a 50 Mb memory. Our contribution is a novel family of models called Budgeted Super Networks (BSN). They are learned using gradient descent techniques applied on a budgeted learning objective function which integrates a maximum authorized cost, while making no assumption on the nature of this cost. We present a set of experiments on computer vision problems and analyze the ability of our technique to deal with three different costs: the computation cost, the memory consumption cost and a distributed computation cost. We particularly show that our model can discover neural network architectures that have a better accuracy than the ResNet and Convolutional Neural Fabrics architectures on CIFAR-10 and CIFAR-100, at a lower cost.