Verfahren und Systeme zur Durchführung einer Affinen Transformation pro Kanal unter Verwendung eines Beschleunigers eines Neuronalen Netzwerks

Details

Veröffentlichungs-Nr.

EP4361892

Aktenzeichen

EP23207084

Anmeldetag

31. Oktober 2023

Veröffentlichung

1. Mai 2024

Rechtsraum

EP

IPC

G06N3/063G06N3/0495// G06N3/0464

Abstract

Methods and neural network accelerators for implementing per channel quantised affine transformation on an input tensor to generate an output tensor, the affine transformation for a channel comprising a multiplication by a multiplication parameter for the channel followed by an addition of an addition parameter for the channel. The neural network accelerator comprises a depth-wise convolution processing unit configured to accelerate depth-wise convolution operations, and a high-precision re-quantisation processing unit configured to accelerate re-quantisation operations. The methods include implementing each affine transformation as an addition followed by a multiplication by: implementing the additions by performing, using the depth-wise convolution processing unit, a 1x1 depth-wise convolution on the input tensor based on a weight and a bias for each channel of the input tensor, wherein the weight for a channel is set to an addition scale factor for the channel and the bias for a channel is set to an integer addition value for the channel; and implementing the multiplications by scaling, using the high-precision re-quantisation processing unit, each value output from the depth-wise convolution processing unit, by a high-precision multiplication value for the corresponding channel.

Anmelder

Firma

Imagination Technologies Limited

Land

🇬🇧 Großbritannien

Fachgebiete