Unlock instant, AI-driven research and patent intelligence for your innovation.

Neural network model training method and device, storage medium and equipment

A technology of neural network model and training method, applied in the direction of biological neural network model, neural learning method, etc., which can solve the problems of inaccurate noise label processing and limited improvement of model anti-interference effect, achieving high accuracy and good explainability performance, good anti-jamming effect

Pending Publication Date: 2020-11-06
深圳智峪生物科技有限公司
View PDF0 Cites 3 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] Embodiments of the present invention provide a training method, device, storage medium and equipment for a neural network model, aiming to solve the technical problem that the existing processing method for noise labels is inaccurate, resulting in limited improvement of the anti-interference effect of the model

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
  • Neural network model training method and device, storage medium and equipment
  • Neural network model training method and device, storage medium and equipment
  • Neural network model training method and device, storage medium and equipment

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0041] see figure 1 , which shows the training method of the neural network model in Embodiment 1 of the present invention, the method specifically includes steps S01-step S03:

[0042] Step S01, acquiring an original data set, and training an original neural network model according to the original data set.

[0043] Wherein, the original data set may be an image sample set, such as a medical image sample set. After the original data set is obtained, the neural network training can be performed based on the original data set to obtain the original neural network model. For example, the image segmentation neural network training can be performed on the image sample set to obtain the image segmentation network model. As another example, the original data set can also be a text data set, and the text data set is trained to obtain a semi-supervised multi-label learning model.

[0044] Step S02, identifying noise labels from the original neural network model.

[0045] During spe...

Embodiment 2

[0056] see figure 2 , shows the training method of the neural network model in the second embodiment of the present invention. The difference between the detection method in this embodiment and the detection method in the first embodiment is that the training method of the neural network model in this embodiment Also further include step S11-step S14:

[0057] Step S11, obtaining an original image sample set, and training an original image segmentation network model according to the original image sample set.

[0058] In this embodiment, the method adopts a teacher-student framework when training the neural network model, and the original image segmentation network model is a teacher model.

[0059] Step S12, using confidence learning technology to identify noise labels from the original image segmentation network model.

[0060] Specifically, step S12 specifically includes the following refinement steps:

[0061] Calculating the predicted probability of the original neura...

Embodiment 3

[0084] On the other hand, the present invention also proposes a training device for a neural network model, please refer to Figure 4 , which shows the training device of the neural network model provided by the third embodiment of the present invention, the device includes:

[0085] Data acquisition module 11, is used for obtaining original data set, and trains original neural network model according to described original data set;

[0086] Noise identification module 12, for identifying noise label from described original neural network model;

[0087] The model training module 13 is used to modify the original data corresponding to the noise label, and train a new neural network model according to the modified data set.

[0088] Wherein, the original data set may be an image sample set, such as a medical image sample set. After the original data set is obtained, the neural network training can be performed based on the original data set to obtain the original neural netwo...

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 is suitable for the technical field of model training, and provides a neural network model training method, device and system. The method comprises the steps of obtaining an original data set, and training an original neural network model according to the original data set; identifying a noise label from the original neural network model; and modifying the noise label, and training anew neural network model according to the modified data set. According to the invention, an original neural network model is trained by using an original data set, a noise label in the original neural network model is identified, an error label in the original data set is determined; after the error label is corrected, the new neural network model is finally trained according to the modified dataset, and the error label is directly determined from the network model and corrected, so accuracy is high, interpretability is very good, and the finally trained new neural network model has a relatively good anti-interference effect.

Description

technical field [0001] The invention belongs to the technical field of model training, and in particular relates to a training method, device, storage medium and equipment of a neural network model. Background technique [0002] Deep learning technology has achieved great success in the field of image processing, and their success is inseparable from the training of neural network models. In the process of training a neural network model, data and corresponding labels (gold standard) are the most critical factors besides the network model. [0003] If there is some noise in the label of the data, that is, the wrong label, it will have a great negative impact on the training of the network, which will lead to the deterioration of the performance of the neural network model, that is, the model is susceptible to interference when the label is polluted by noise. Therefore, how to ensure the performance of the network model in the presence of noise in the label, so that the trai...

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
IPC IPC(8): G06N3/08
CPCG06N3/08
Inventor 李镇张敏清
Owner 深圳智峪生物科技有限公司