http://rdf.ncbi.nlm.nih.gov/pubchem/patent/CN-107346328-B
Outgoing Links
Predicate | Object |
---|---|
classificationCPCInventive | http://rdf.ncbi.nlm.nih.gov/pubchem/patentcpc/G06F16-43 http://rdf.ncbi.nlm.nih.gov/pubchem/patentcpc/G06N3-045 http://rdf.ncbi.nlm.nih.gov/pubchem/patentcpc/G06N3-08 |
classificationIPCInventive | http://rdf.ncbi.nlm.nih.gov/pubchem/patentipc/G06N3-08 http://rdf.ncbi.nlm.nih.gov/pubchem/patentipc/G06F16-43 http://rdf.ncbi.nlm.nih.gov/pubchem/patentipc/G06N3-04 |
filingDate | 2017-05-25-04:00^^<http://www.w3.org/2001/XMLSchema#date> |
grantDate | 2020-09-08-04:00^^<http://www.w3.org/2001/XMLSchema#date> |
publicationDate | 2020-09-08-04:00^^<http://www.w3.org/2001/XMLSchema#date> |
publicationNumber | CN-107346328-B |
titleOfInvention | A Cross-modal Association Learning Method Based on Multi-granularity Hierarchical Networks |
abstract | The present invention relates to a cross-modal association learning method based on a multi-granularity hierarchical network, comprising the following steps: 1. establishing a cross-modal database including multiple modal types, and dividing the data in the database into training sets, verification Set and test set, perform block processing on different modal data in the database, and extract all modal original data and feature vectors of the block data. 2. Use the original data and the chunked data to train a multi-granularity hierarchical network structure to learn a unified representation for different modal data. 3. Using the trained multi-granularity hierarchical network structure, a unified representation of different modal data is obtained, and then the similarity of different modal data is calculated. 4. Take any modal type in the test set as the query modal, take another modal type as the target modal, calculate the similarity between the query sample and the query target, and obtain the correlation of the target modal data according to the similarity. list of results. The present invention can improve the accuracy of cross-modal retrieval. |
priorityDate | 2017-05-25-04:00^^<http://www.w3.org/2001/XMLSchema#date> |
type | http://data.epo.org/linked-data/def/patent/Publication |
Incoming Links
Total number of triples: 18.