Image classification systems based on CNN based IC and light-weight classifier

Patent metadata

Field	Value
Publication number	US-10402628-B2
Application number	US-201815963990-A
Country	US
Kind code	B2
Filing date	Apr 26, 2018
Priority date	Oct 10, 2016
Publication date	Sep 3, 2019
Grant date	Sep 3, 2019

Abstract

Image classification system contains a CNN based IC configured for extracting features out of input data by performing convolution operations using filter coefficients of ordered convolutional layers and a classifier IC configured for classifying the input data using reduced set of the extracted features based on a light-weight classifier. Light-weight classifier is derived by: training filter coefficients of the ordered convolutional layers using a dataset containing N labeled data, the trained filter coefficients are for the CNN based IC; outputting respective extracted features of the N labeled data after performing convolution operations of ordered convolutional layers using the trained filter coefficients, each labeled data contains X features; creating the reduced set of the extracted features by eliminating those of the X features that contain zeros in at least M of the N labeled data; and adjusting M until the light-weight classifier achieves satisfactory results using the reduced set.

First claim

What is claimed is:

1. An image classification system comprising: a cellular neural networks (CNN) based integrated circuit (IC) configured for extracting features out of an input data by performing convolution operations using filter coefficients of a plurality of ordered convolutional layers, the CNN based IC comprising a plurality of cellular neural networks (CNN) processing engines operatively coupled to at least one input/output data bus, the plurality of CNN processing engines being connected in a loop with a clock-skew circuit, each CNN processing engine comprising: a CNN processing block configured for simultaneously obtaining results of the convolution operations; a first set of memory buffers operatively coupling to the CNN processing block for storing the input data; and a second set of memory buffers operative coupling to the CNN processing block for storing the filter coefficients; and a classifier IC configured for classifying the input data into a set of predefined categories using a reduced set of the extracted features based on a light-weight classifier derived by a method containing following operations: training the filter coefficients of the ordered convolutional layers using a dataset containing N labeled data in a deep learning model for image classification, the trained filter coefficients being configured for the CNN based IC; outputting respective feature vectors of the N labeled data after performing convolution operations of the ordered convolutional layers using the trained filter coefficients stored in the CNN based IC, each of the N labeled data containing X extracted features in the corresponding feature vector; creating the reduced set of the extracted features by eliminating those of the X extracted features that contain zeros in at least M of the N labeled data; and iteratively deriving the light-weight classifier using the reduced set of the extracted features by adjusting M until the light-weight classifier achieves results in accordance with image classification criteria, where M, N and X are positive integers.

2. The image classification system of claim 1 , wherein the classifier IC comprising logic circuits for the light-weight classifier based on decision tree.

3. The image classification system of claim 1 , wherein the classifier IC comprising logic circuits for the light-weight classifier based on logistic regression.

4. The image classification system of claim 1 , wherein the deep learning model comprises Visual Geometry Group's VGG16 model with 13 ordered convolutional layers and 3 Fully-Connected layers.

5. The image classification system of claim 1 , wherein the deep learning model comprises Visual Geometry Group's VGG16 model with 13 ordered convolutional layers and 3 Fully-Connected layers.

6. The image classification system of claim 1 , wherein each of the filter coefficients comprises bi-valued 3×3 filter kernel.

7. The image classification system of claim 1 , wherein the CNN based IC and the classifier IC are coupled to each other via a network bus.

8. The image classification system of claim 7 , wherein the network bus comprises a Universal Serial Bus.

9. The image classification system of claim 7 , wherein the network bus comprises a Peripheral Component Interconnect Express bus.

10. The image classification system of claim 1 , wherein the CNN based IC is further configured for performing activation and pooling operations.

Assignees

• Gyrfalcon Tech Inc

Inventors

• Yang Lin
• Dong Patrick Z
• Young Charles Jin
• Dong Jason
• Zhang Wenhan
• Sun Baohua

Classifications

• G06V30/18057

  Integrating the filters into a hierarchical structure, e.g. convolutional neural networks [CNN] · CPC title

• G06V30/19173

  Classification techniques · CPC title

• G06V10/82 Primary

  using neural networks · CPC title

• G06V40/172

  Classification, e.g. identification · CPC title

• G06N3/045

  Combinations of networks · CPC title