http://rdf.ncbi.nlm.nih.gov/pubchem/patent/CN-105100894-B
Outgoing Links
Predicate | Object |
---|---|
classificationCPCAdditional | http://rdf.ncbi.nlm.nih.gov/pubchem/patentcpc/G06T2207-30201 http://rdf.ncbi.nlm.nih.gov/pubchem/patentcpc/G06T2207-10016 |
classificationCPCInventive | http://rdf.ncbi.nlm.nih.gov/pubchem/patentcpc/G06V20-49 http://rdf.ncbi.nlm.nih.gov/pubchem/patentcpc/G06T5-002 http://rdf.ncbi.nlm.nih.gov/pubchem/patentcpc/G06F18-2411 http://rdf.ncbi.nlm.nih.gov/pubchem/patentcpc/H04N7-18 http://rdf.ncbi.nlm.nih.gov/pubchem/patentcpc/G06V40-161 http://rdf.ncbi.nlm.nih.gov/pubchem/patentcpc/G06F16-951 http://rdf.ncbi.nlm.nih.gov/pubchem/patentcpc/G06V20-41 http://rdf.ncbi.nlm.nih.gov/pubchem/patentcpc/H04N21-44008 http://rdf.ncbi.nlm.nih.gov/pubchem/patentcpc/G06V20-30 http://rdf.ncbi.nlm.nih.gov/pubchem/patentcpc/G06F18-2155 http://rdf.ncbi.nlm.nih.gov/pubchem/patentcpc/G06F40-169 http://rdf.ncbi.nlm.nih.gov/pubchem/patentcpc/G06F16-21 http://rdf.ncbi.nlm.nih.gov/pubchem/patentcpc/G06V40-172 http://rdf.ncbi.nlm.nih.gov/pubchem/patentcpc/G06F16-5846 |
classificationIPCInventive | http://rdf.ncbi.nlm.nih.gov/pubchem/patentipc/H04N21-44 http://rdf.ncbi.nlm.nih.gov/pubchem/patentipc/G06K9-00 http://rdf.ncbi.nlm.nih.gov/pubchem/patentipc/H04N7-18 http://rdf.ncbi.nlm.nih.gov/pubchem/patentipc/G06K9-62 |
filingDate | 2015-08-24-04:00^^<http://www.w3.org/2001/XMLSchema#date> |
grantDate | 2020-05-05-04:00^^<http://www.w3.org/2001/XMLSchema#date> |
publicationDate | 2020-05-05-04:00^^<http://www.w3.org/2001/XMLSchema#date> |
publicationNumber | CN-105100894-B |
titleOfInvention | Face automatic labeling method and system |
abstract | The invention discloses a face automatic labeling method and system. Wherein, the method includes: after dividing the input video into different video frame sets, extracting temporal and spatial information from the video frame set by acquiring content from a camera and using a shot boundary detection algorithm. And collect weakly labeled data by scraping weakly labeled facial images in social networks. Then, the noise in the weakly labeled data is filtered out in combination with the face detection of the iteratively optimized clustering algorithm, and a labeled database containing the optimized labeled images is generated as training data. Based on the optimized marker images stored in the marker database, a target video frame containing a face image that matches any of the optimized marker images in the marker database is found and marked in the input video. Through a semi-supervised learning algorithm, the unlabeled facial trajectories in the input video are labeled to complete the video face annotation. Finally, the output contains an input video with annotated face images. |
priorityDate | 2014-08-26-04:00^^<http://www.w3.org/2001/XMLSchema#date> |
type | http://data.epo.org/linked-data/def/patent/Publication |
Incoming Links
Total number of triples: 32.