Small sample online training method and device and storage medium
A training method and small sample technology, applied in neural learning methods, neural architectures, biological neural network models, etc., can solve the problem of difficult to carry mobile devices and embedded platforms, limited storage, power consumption and computing power, and complex computing. Training parameters and other issues to achieve the effect of solving catastrophic forgetting, portability and energy efficiency, and high training accuracy
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment 1
[0050] This embodiment proposes a small sample online training method, and the method is implemented through the following technical solutions:
[0051] By using the CIFAR100 data set to train the VGG16 neural network, the weight parameters are obtained, and the weight parameters are used as the original training parameters of the pre-training model;
[0052] Collect and mark samples in actual application scenarios. It should be noted that the specified preset amount of collected samples batch_size=32, and when the specified preset amount of collected samples batch_size=32, input training is performed on the samples;
[0053] Input the labeled samples and original training parameters into the pre-training model for several training iterations to update the original training parameters;
[0054]The pre-training model is obtained based on the online training neural network. The online training neural network includes convolutional layer, interval batch normalization module, RELU...
Embodiment 2
[0110] A small sample online training device provided by an embodiment of the present invention includes a memory and a processor;
[0111] the storage medium is used to store instructions;
[0112] The processor is configured to operate in accordance with the instructions to perform the steps of the following methods:
[0113] The neural network is trained through the data set to obtain the weight parameters, and the weight parameters are used as the original training parameters of the pre-training model;
[0114] Collect and label a specified preset amount of samples;
[0115] The labeled samples and original training parameters are input into the pre-trained model for several training iterations to update the original training parameters.
Embodiment 3
[0117] The computer-readable storage medium provided by the embodiment of the present invention stores a computer program on it, and when the program is executed by a processor, implements the steps of the following method:
[0118] The neural network is trained through the data set to obtain the weight parameters, and the weight parameters are used as the original training parameters of the pre-training model;
[0119] Collect and label a specified preset amount of samples;
[0120] The labeled samples and original training parameters are input into the pre-trained model for several training iterations to update the original training parameters.
PUM
Abstract
Description
Claims
Application Information
- R&D Engineer
- R&D Manager
- IP Professional
- Industry Leading Data Capabilities
- Powerful AI technology
- Patent DNA Extraction
Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.
© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com