1D-FALCON: Accelerating Deep Convolutional Neural Network Inference by Co-optimization of Models and Underlying Arithmetic Implementation

Partha Maji, Robert D. Mullins

Abstract: Deep convolutional neural networks (CNNs), which are at the heart of many new emerging applications, achieve remarkable performance in audio and visual recognition tasks, at the expense of high computational complexity, limiting their deployability. In modern CNNs it is typical for the convolution layers to consume the vast majority of the compute resources during inference. This has made the acceleration of these layers an important research and industrial goal. In this paper, we examine the effects of co-optimizing the internal structures of the convolutional layers and underlying implementation of fundamental convolution operation. We demonstrate that a combination of these methods can have a big impact on the overall speed-up of a CNN, achieving a tenfold increase over baseline. We also introduce a new class of fast 1-D convolutions for CNNs using the Toom-Cook algorithm. We show that our proposed scheme is mathematically well grounded, robust, does not require any time-consuming retraining, and still achieves speed-ups solely from convolutional layers with no loss in baseline accuracy.