Matches in SemOpenAlex for { <https://semopenalex.org/work/W4386076031> ?p ?o ?g. }
Showing items 1 to 88 of
88
with 100 items per page.
- W4386076031 abstract "Recent years have witnessed significant developments in point cloud processing, including classification and segmentation. However, supervised learning approaches need a lot of well-labeled data for training, and annotation is labor-and time-intensive. Self-supervised learning, on the other hand, uses unlabeled data, and pretrains a back-bone with a pretext task to extract latent representations to be used with the downstream tasks. Compared to 2D images, self-supervised learning of 3D point clouds is under-explored. Existing models, for self-supervised learning of 3D point clouds, rely on a large number of data samples, and require significant amount of computational re-sources and training time. To address this issue, we propose a novel contrastive learning approach, referred to as To ThePoint. Different from traditional contrastive learning methods, which maximize agreement between features obtained from a pair of point clouds formed only with different types of augmentation, ToThePoint also maximizes the agreement between the permutation invariant features and features discarded after max pooling. We first perform self-supervised learning on the ShapeNet dataset, and then evaluate the performance of the network on different downstream tasks. In the downstream task experiments, performed on the ModelNet40, ModelNet40C, ScanobjectNN and ShapeNet-Part datasets, our proposed ToThe-Point achieves competitive, if not better results compared to the state-of-the-art baselines, and does so with significantly less training time (200 times faster than baselines)." @default.
- W4386076031 created "2023-08-23" @default.
- W4386076031 creator A5004337702 @default.
- W4386076031 creator A5030992210 @default.
- W4386076031 creator A5042931416 @default.
- W4386076031 creator A5050119204 @default.
- W4386076031 creator A5061391892 @default.
- W4386076031 creator A5068697970 @default.
- W4386076031 date "2023-06-01" @default.
- W4386076031 modified "2023-09-27" @default.
- W4386076031 title "ToThePoint: Efficient Contrastive Learning of 3D Point Clouds via Recycling" @default.
- W4386076031 cites W1644641054 @default.
- W4386076031 cites W2211722331 @default.
- W4386076031 cites W2553307952 @default.
- W4386076031 cites W2927438889 @default.
- W4386076031 cites W2963231572 @default.
- W4386076031 cites W2963719584 @default.
- W4386076031 cites W2964228567 @default.
- W4386076031 cites W2979750740 @default.
- W4386076031 cites W2981440248 @default.
- W4386076031 cites W2990500698 @default.
- W4386076031 cites W3034266912 @default.
- W4386076031 cites W3035524453 @default.
- W4386076031 cites W3039448353 @default.
- W4386076031 cites W3116959466 @default.
- W4386076031 cites W3119708198 @default.
- W4386076031 cites W3128661784 @default.
- W4386076031 cites W3141954417 @default.
- W4386076031 cites W3167071962 @default.
- W4386076031 cites W4239510810 @default.
- W4386076031 cites W4312317653 @default.
- W4386076031 cites W4312697367 @default.
- W4386076031 doi "https://doi.org/10.1109/cvpr52729.2023.02086" @default.
- W4386076031 hasPublicationYear "2023" @default.
- W4386076031 type Work @default.
- W4386076031 citedByCount "0" @default.
- W4386076031 crossrefType "proceedings-article" @default.
- W4386076031 hasAuthorship W4386076031A5004337702 @default.
- W4386076031 hasAuthorship W4386076031A5030992210 @default.
- W4386076031 hasAuthorship W4386076031A5042931416 @default.
- W4386076031 hasAuthorship W4386076031A5050119204 @default.
- W4386076031 hasAuthorship W4386076031A5061391892 @default.
- W4386076031 hasAuthorship W4386076031A5068697970 @default.
- W4386076031 hasConcept C119857082 @default.
- W4386076031 hasConcept C131979681 @default.
- W4386076031 hasConcept C154945302 @default.
- W4386076031 hasConcept C162324750 @default.
- W4386076031 hasConcept C187736073 @default.
- W4386076031 hasConcept C2524010 @default.
- W4386076031 hasConcept C2780451532 @default.
- W4386076031 hasConcept C28006648 @default.
- W4386076031 hasConcept C28719098 @default.
- W4386076031 hasConcept C33923547 @default.
- W4386076031 hasConcept C41008148 @default.
- W4386076031 hasConcept C58973888 @default.
- W4386076031 hasConcept C70437156 @default.
- W4386076031 hasConcept C89600930 @default.
- W4386076031 hasConceptScore W4386076031C119857082 @default.
- W4386076031 hasConceptScore W4386076031C131979681 @default.
- W4386076031 hasConceptScore W4386076031C154945302 @default.
- W4386076031 hasConceptScore W4386076031C162324750 @default.
- W4386076031 hasConceptScore W4386076031C187736073 @default.
- W4386076031 hasConceptScore W4386076031C2524010 @default.
- W4386076031 hasConceptScore W4386076031C2780451532 @default.
- W4386076031 hasConceptScore W4386076031C28006648 @default.
- W4386076031 hasConceptScore W4386076031C28719098 @default.
- W4386076031 hasConceptScore W4386076031C33923547 @default.
- W4386076031 hasConceptScore W4386076031C41008148 @default.
- W4386076031 hasConceptScore W4386076031C58973888 @default.
- W4386076031 hasConceptScore W4386076031C70437156 @default.
- W4386076031 hasConceptScore W4386076031C89600930 @default.
- W4386076031 hasFunder F4320321001 @default.
- W4386076031 hasLocation W43860760311 @default.
- W4386076031 hasOpenAccess W4386076031 @default.
- W4386076031 hasPrimaryLocation W43860760311 @default.
- W4386076031 hasRelatedWork W2597787948 @default.
- W4386076031 hasRelatedWork W2917062864 @default.
- W4386076031 hasRelatedWork W2959771705 @default.
- W4386076031 hasRelatedWork W3015465855 @default.
- W4386076031 hasRelatedWork W3093454656 @default.
- W4386076031 hasRelatedWork W3211762283 @default.
- W4386076031 hasRelatedWork W4311005618 @default.
- W4386076031 hasRelatedWork W4319309271 @default.
- W4386076031 hasRelatedWork W4320063314 @default.
- W4386076031 hasRelatedWork W4323060038 @default.
- W4386076031 isParatext "false" @default.
- W4386076031 isRetracted "false" @default.
- W4386076031 workType "article" @default.