Hash model training method, similar object retrieval method and device
A model training, object technology, applied in the field of computer vision
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment Construction
[0054] The goal of traditional unsupervised hashing methods is to achieve distance preservation between the original feature space (usually Euclidean space) and Hamming space. In order to enable the unsupervised hash method to obtain higher approximate nearest neighbor retrieval accuracy on large-scale data sets, the present invention provides a hash model training method, which solves the problem of accuracy from the perspective of neighborhood preservation, Specifically, k neighbor points of any data (ie, reference point) in the training data set are used as the neighborhood. Among them, the present invention builds a multi-layer neighborhood pyramid by gradually increasing the number of neighbor points. In each layer of the neighborhood pyramid, the average distance from the neighbor points in the layer to the reference point is taken as the Euclidean neighborhood measure of the layer's neighborhood. The data points in the original space are mapped to the Hamming space usi...
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