Combined query image retrieval method based on multi-order adversarial feature learning
A technology of feature learning and combined query, applied in still image data retrieval, metadata still image retrieval, digital data information retrieval, etc., can solve the problems of insufficient use of multi-scale and low retrieval efficiency
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment Construction
[0045]The present invention will be described in detail below with reference to the accompanying drawings and specific embodiments.
[0046]The present invention proposes a combination of multi-order counterfeit characteristics, including the following steps:
[0047](1) Extraction of the reference image and the characteristics of the target image and the characteristics of the target image are extracted by different feature extraction methods, and the initial features of these two modal data are obtained.
[0048](1-1) Reference image of given inputsAnd target image Xt, Use the MobileNet or RESNet18 network model extracted by the imagenet data set to extract image featuresWhere i represents the level of the network, the actual model is extracted from low, medium and high-level.
[0049](1-2) Modifying text T of the given input T, first use simple vocabulary and embedded layers to convert words to the word embed vector {w1W2, ... wn}, Where WnIndicate the embedded vector of the nth word. With t...
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