Convolutional Neural Network Quantization using Generalized Gamma Distribution
This work addresses the need for efficient CNN deployment on edge devices, offering incremental improvements in quantization for hardware accelerators.
The authors tackled the problem of deploying convolutional neural networks (CNNs) on energy-efficient hardware accelerators by proposing a novel quantization algorithm that minimizes quantization error for weights and biases and uses a generalized gamma distribution to optimize feature-map quantization, achieving a higher signal-to-quantization-noise ratio (SQNR) than previous methods.
As edge applications using convolutional neural networks (CNN) models grow, it is becoming necessary to introduce dedicated hardware accelerators in which network parameters and feature-map data are represented with limited precision. In this paper we propose a novel quantization algorithm for energy-efficient deployment of the hardware accelerators. For weights and biases, the optimal bit length of the fractional part is determined so that the quantization error is minimized over their distribution. For feature-map data, meanwhile, their sample distribution is well approximated with the generalized gamma distribution (GGD), and accordingly the optimal quantization step size can be obtained through the asymptotical closed form solution of GGD. The proposed quantization algorithm has a higher signal-to-quantization-noise ratio (SQNR) than other quantization schemes previously proposed for CNNs, and even can be more improved by tuning the quantization parameters, resulting in efficient implementation of the hardware accelerators for CNNs in terms of power consumption and memory bandwidth.