Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Face noise data set CNN training method based on overall cosine distribution

A data set and training data set technology, applied in the field of image recognition, can solve problems such as difficult acquisition of prior knowledge, large fluctuations in loss values, poor recognition effect, etc., and achieve small memory resources, large representation benefits, and small computing resources. Effect

Active Publication Date: 2019-09-17
BEIJING YUNSHITU INFORMATION TECH CO LTD
View PDF2 Cites 5 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0013] 2.2) The dependence on hyperparameters is serious, and the convergence of the training process is difficult;
[0014] 2.3) For symmetrical intra-class noise, the recognition effect is relatively poor;
[0017] 3.1) This method needs to pre-determine the noise ratio of the data set. Usually, this prior knowledge is difficult to obtain, thus limiting its usage scenarios;
[0018] 3.2) Since the loss value of a single small batch sample set fluctuates greatly, it is not very accurate to filter noise data based on the small loss of a single small batch sample set;
[0019] 3.3) This method requires an isomorphic dual network structure, and the video memory usage and calculation consumption are basically twice that of the general network, which is not very practical for large networks with limited video memory;

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Face noise data set CNN training method based on overall cosine distribution
  • Face noise data set CNN training method based on overall cosine distribution
  • Face noise data set CNN training method based on overall cosine distribution

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0068] The present invention proposes a face noise data set CNN training method based on the overall cosine distribution, which will be further described in detail below in conjunction with the accompanying drawings and specific embodiments. The embodiments described in the present invention are exemplary, and are only used for explaining the present invention, but not construed as limiting the present invention.

[0069] The present invention proposes a kind of face noise data set CNN training method based on overall cosine distribution, comprises the following steps:

[0070] 1) Select the face training data set containing noise and record it as D all (the human face training data set comprises human face sample picture and the corresponding label of each picture, can adopt off-the-shelf human face training data set, wherein the number of label categories is not less than 1000, the number of human face sample pictures of each category Not less than 10), construct a benchmar...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention provides a face noise data set CNN training method based on overall cosine distribution, and belongs to the field of image recognition. The method comprises the following steps: firstly, selecting a noise-containing face training data set and constructing a reference convolutional neural network trunk; adding layers to the convolutional neural network trunk to obtain an updated convolutional neural network; sequentially obtaining a small-batch sample set from the face training data set, inputting each face image sample of the small-batch sample set into the updated network to obtain a corresponding cosine value, and adding the cosine value into cosine distribution; and calculating a corresponding training weight for each cosine value by designing three strategies and fusion functions, learning the updated network by using a gradient descent algorithm, and ending the training when the number of iterations reaches an upper limit. The method has good robustness, and can quickly and efficiently train a face training data set containing noise under the condition that priori knowledge is not needed.

Description

technical field [0001] The invention relates to the field of image recognition. In particular, it relates to a face noise dataset CNN training method based on the overall cosine distribution. Background technique [0002] In recent years, deep learning technology has made rapid progress in both algorithms and hardware, and the scale of data it relies on has become larger and larger. Taking face recognition as an example, training a convolutional neural network for face recognition CNN requires large-scale training data sets. Although most of these training data sets are cleaned and filtered manually or by machines, there is still a certain proportion of noise. Face training data refers to face sample pictures and their corresponding labels, and face noise data refers to face sample pictures marked with labels that are not real labels. The IMDB-Face method mentioned that Microsoft's MS-Celeb-1M training data set contains nearly 50% noise data, and the depth model trained di...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Applications(China)
IPC IPC(8): G06K9/00
CPCG06V40/172
Inventor 黄杨昱胡伟袁国栋
Owner BEIJING YUNSHITU INFORMATION TECH CO LTD
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products