Method for Hybrid Precision Convolutional Neural Network Representation

TL;DR

This paper introduces a novel method for representing CNNs using hybrid precision to optimize power efficiency and throughput without sacrificing accuracy in integrated circuit implementations.

Contribution

It proposes a hybrid precision representation approach for CNNs that balances accuracy and efficiency, addressing limitations of uniform fixed-point quantization.

Findings

Improved power efficiency in CNN implementations.

Enhanced throughput with maintained accuracy.

Effective trade-offs between precision and performance.

Abstract

This invention addresses fixed-point representations of convolutional neural networks (CNN) in integrated circuits. When quantizing a CNN for a practical implementation there is a trade-off between the precision used for operations between coefficients and data and the accuracy of the system. A homogenous representation may not be sufficient to achieve the best level of performance at a reasonable cost in implementation complexity or power consumption. Parsimonious ways of representing data and coefficients are needed to improve power efficiency and throughput while maintaining accuracy of a CNN.