Predicate |
Object |
assignee |
http://rdf.ncbi.nlm.nih.gov/pubchem/patentassignee/MD5_d2d422b9357e5be60f36faf00a3ae5be |
classificationCPCInventive |
http://rdf.ncbi.nlm.nih.gov/pubchem/patentcpc/G06N3-084 http://rdf.ncbi.nlm.nih.gov/pubchem/patentcpc/G06N3-045 http://rdf.ncbi.nlm.nih.gov/pubchem/patentcpc/G06T17-00 http://rdf.ncbi.nlm.nih.gov/pubchem/patentcpc/G06F18-24147 http://rdf.ncbi.nlm.nih.gov/pubchem/patentcpc/G06N3-088 |
classificationIPCInventive |
http://rdf.ncbi.nlm.nih.gov/pubchem/patentipc/G06N3-04 http://rdf.ncbi.nlm.nih.gov/pubchem/patentipc/G06N3-08 http://rdf.ncbi.nlm.nih.gov/pubchem/patentipc/G06K9-62 http://rdf.ncbi.nlm.nih.gov/pubchem/patentipc/G06T17-00 |
filingDate |
2021-11-30-04:00^^<http://www.w3.org/2001/XMLSchema#date> |
inventor |
http://rdf.ncbi.nlm.nih.gov/pubchem/patentinventor/MD5_da432347c736cef86ce98c80f11d23c5 http://rdf.ncbi.nlm.nih.gov/pubchem/patentinventor/MD5_8e27dd75d1a99b9be5473cb272ac1ec3 http://rdf.ncbi.nlm.nih.gov/pubchem/patentinventor/MD5_174394d927b9dee738e58f124e3e4dfe http://rdf.ncbi.nlm.nih.gov/pubchem/patentinventor/MD5_e543e43b1e2d3a517bcc9d1f19aecf81 http://rdf.ncbi.nlm.nih.gov/pubchem/patentinventor/MD5_0d80b7e5c42d2d40714e05a4945c3238 |
publicationDate |
2022-02-25-04:00^^<http://www.w3.org/2001/XMLSchema#date> |
publicationNumber |
CN-114092650-A |
titleOfInvention |
A 3D point cloud generation method based on efficient graph convolution |
abstract |
The invention discloses a three-dimensional point cloud generation method based on high-efficiency graph convolution, and relates to the technical field of deep learning and three-dimensional point cloud data enhancement. Taking the three-dimensional point cloud generation as the goal, the used network model is added on the basis of a fully connected layer. The graph convolution model is used to improve the point cloud feature learning ability of the generator network, and the overall network structure is optimized by using the idea of feature fusion. The redundant calculation is reduced by sharing the calculation results of KNN, and the order of feature aggregation and MLP is adjusted. The graph convolution structure is improved, which reduces the computational complexity and improves the feature learning ability of the network. |
priorityDate |
2021-11-30-04:00^^<http://www.w3.org/2001/XMLSchema#date> |
type |
http://data.epo.org/linked-data/def/patent/Publication |