Generative adversarial migration learning method based on sketch annotation information
A technology of labeling information and transfer learning, applied in the field of cross-domain image classification, which can solve the problem of inability to judge the invariant features of the feature domain.
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Example Embodiment
[0073] The present invention provides a method for generating anti-migration learning based on sketch labeling information, obtaining an initial sketch map, and a paired data set constructed in the form of "source image-source image edge labeling map"; constructing an edge based on sketch labeling information Segment deep networks and train them; select target domain samples based on matrix norm; construct and train a generation of anti-migration learning network based on sketch annotation information, which includes depth generator network, depth discriminator network, and edge segmentation depth based on sketch annotation information Network and deep classifier network; input the target domain image to obtain the classification result of the target domain image; the present invention uses the similarity of the source domain data and the target domain data structure to generate a sample with a certain label conforming to the target domain distribution through structural constrai...
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.
© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap