Unlock instant, AI-driven research and patent intelligence for your innovation.
A neural network training method and device
What is Al technical title?
Al technical title is built by PatSnap Al team. It summarizes the technical point description of the patent document.
A neural network and neural network model technology, applied in neural learning methods, biological neural network models, instruments, etc., can solve the problems that the training image set is not well applicable and the accuracy needs to be improved, so as to reduce labor costs and improve The effect of accuracy
Active Publication Date: 2021-04-02
BEIJING TUSEN ZHITU TECH CO LTD
View PDF10 Cites 0 Cited by
Summary
Abstract
Description
Claims
Application Information
AI Technical Summary
This helps you quickly interpret patents by identifying the three key elements:
Problems solved by technology
Method used
Benefits of technology
Problems solved by technology
[0003] Based on the existing training image set, the trained neural network is only suitable for the situation that appears in the training image set, and it is not suitable for the situation that does not appear in the training image set, and the accuracy needs to be improved.
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
Click on the blue label to locate the original text in one second.
Reading with bidirectional positioning of images and text.
Smart Image
Examples
Experimental program
Comparison scheme
Effect test
Embodiment 1
[0023] see figure 1 , which is a flowchart of a neural network training method in an embodiment of the present invention, the method comprising:
[0024] Step 101, using the training image set to train the neural network model to obtain a primary neural network.
[0025] Step 102: Input the test images in the test image set into the primary neural network, and output the detection results corresponding to the test images.
[0026] Step 103 , selecting test images with incorrect detection results and adding them to the training image set to obtain a new training image set, and performing step 104 .
[0027] Step 104, using a new training image set to train the primary neural network.
[0028] Preferably, in order to further improve the robustness and accuracy of the trained neural network, in the embodiment of the present invention, after step 102 outputs the detection result corresponding to the test image, before step 103, it is judged whether the current primary neural net...
Embodiment 2
[0053] Based on the same idea as the neural network training method provided in the first embodiment, the second embodiment of the present invention provides a neural network training device, the structure of which is as follows image 3 shown, including:
[0054] The first training unit 31 is used to train the neural network model by using the training image set to obtain the primary neural network;
[0055] The test unit 32 is used to input the test image in the test image set into the primary neural network, and output the detection result corresponding to the test image;
[0056] The update unit 33 is used to select the wrong test image and add it to the training image set to obtain a new training image set, and execute the second training unit 34;
[0057] The second training unit 34 is configured to use a new training image set to train the primary neural network.
[0058] Preferably, in the embodiment of the present invention, image 3 The shown device can further in...
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
Login to View More
Abstract
The invention discloses a neural network training method and device to improve the robustness and accuracy of the neural network. The method includes: using the training image set to train the neural network model to obtain the primary neural network; inputting the test image in the test image set into the primary neural network, and outputting the detection result corresponding to the test image; selecting the test image with an incorrect detection result and adding it to the A new training image set is obtained from the training image set; the primary neural network is trained using the new training image set.
Description
technical field [0001] The invention relates to the field of computers, in particular to a neural network training method and device. Background technique [0002] At present, training a neural network is mainly based on a preset training image set to train a neural network model to obtain a target neural network. The more images in the training image set, the higher the accuracy of the target neural network trained based on the training image set. However, in practical applications, the larger the number of training image sets, the greater the workload of manual labeling, and the higher the labor cost. Therefore, in order to balance the labor cost and the accuracy of the neural network, a certain number of images are selected for labeling to form a training image set. [0003] Based on the existing training image set, the trained neural network is only suitable for the situation that appears in the training image set, and it is not suitable for the situation that does not...
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
Application Date:The date an application was filed.
Publication Date:The date a patent or application was officially published.
First Publication Date:The earliest publication date of a patent with the same application number.
Issue Date:Publication date of the patent grant document.
PCT Entry Date:The Entry date of PCT National Phase.
Estimated Expiry Date:The statutory expiry date of a patent right according to the Patent Law, and it is the longest term of protection that the patent right can achieve without the termination of the patent right due to other reasons(Term extension factor has been taken into account ).
Invalid Date:Actual expiry date is based on effective date or publication date of legal transaction data of invalid patent.
Login to View More 
Login to View More