Patent · US Active

Apparatus and methods for training in convolutional neural networks

US10643129B2 · kind B2 · utility

14Cited by
2References
27Claims
0Family size

Assignee

Inventors

Key dates

Filing dateOct 29, 2018
Grant dateMay 5, 2020
Priority date
Expiry dateOct 29, 2038

Classification

  • Technology area (CPC Y)Emerging Cross-Sectional Technologies
  • CPC primaryY02D10/00
  • WIPO fieldComputer technology
  • WIPO sectorElectrical engineering

Abstract

Aspects for backpropagation of a convolutional neural network are described herein. The aspects may include a direct memory access unit configured to receive input data from a storage device and a master computation module configured to select one or more portions of the input data based on a predetermined convolution window. Further, the aspects may include one or more slave computation modules respectively configured to convolute one of the one or more portions of the input data with one of one or more previously calculated first data gradients to generate a kernel gradient, wherein the master computation module is further configured to update a prestored convolution kernel based on the kernel gradient.

Source: USPTO / EPO open patent data. Objective bibliographic and citation counts.