Matches in SemOpenAlex for { <https://semopenalex.org/work/W4386280904> ?p ?o ?g. }
Showing items 1 to 60 of
60
with 100 items per page.
- W4386280904 endingPage "1" @default.
- W4386280904 startingPage "1" @default.
- W4386280904 abstract "Contrastive representation learning purposes to learn rich point cloud representations from unlabeled data. However, it currently lacks well-designed data augmentation, which can lead to poor regularization and generalization in the point cloud domain. Mixup has been shown to be a simple and effective data augmentation in the image domain. Yet, it was found ineffective to directly applying it to point cloud domain. In this brief, we propose a novel mixup strategy based on the cosine similarity algorithm in point cloud contrastive representation learning. First, we adaptively select pairs of samples with high dissimilarity in the mini-batch for mixing. The Principal Component Analysis (PCA) dimensionality reduction algorithm is used to reduce the cost when calculating the similarity. Then we introduce N-pair contrastive loss that fits to the virtual labels assigned for mixing. Our method outperforms other methods in a diverse range of downstream tasks including 3D object classification and few-shot object classification. Our method can be extended to manifold mixup for better regularization. Meanwhile, the ablation experiments demonstrate that our method is an effective data augmentation strategy." @default.
- W4386280904 created "2023-08-31" @default.
- W4386280904 creator A5005663866 @default.
- W4386280904 creator A5050061521 @default.
- W4386280904 creator A5074363493 @default.
- W4386280904 creator A5079623170 @default.
- W4386280904 creator A5080837299 @default.
- W4386280904 date "2023-01-01" @default.
- W4386280904 modified "2023-10-02" @default.
- W4386280904 title "Cosine Mixup: A Mixup Strategy for Point Cloud Contrastive Representaion Learning" @default.
- W4386280904 doi "https://doi.org/10.1109/tcsii.2023.3310192" @default.
- W4386280904 hasPublicationYear "2023" @default.
- W4386280904 type Work @default.
- W4386280904 citedByCount "0" @default.
- W4386280904 crossrefType "journal-article" @default.
- W4386280904 hasAuthorship W4386280904A5005663866 @default.
- W4386280904 hasAuthorship W4386280904A5050061521 @default.
- W4386280904 hasAuthorship W4386280904A5074363493 @default.
- W4386280904 hasAuthorship W4386280904A5079623170 @default.
- W4386280904 hasAuthorship W4386280904A5080837299 @default.
- W4386280904 hasConcept C103278499 @default.
- W4386280904 hasConcept C111919701 @default.
- W4386280904 hasConcept C11413529 @default.
- W4386280904 hasConcept C115961682 @default.
- W4386280904 hasConcept C131979681 @default.
- W4386280904 hasConcept C153180895 @default.
- W4386280904 hasConcept C154945302 @default.
- W4386280904 hasConcept C2776135515 @default.
- W4386280904 hasConcept C2780762811 @default.
- W4386280904 hasConcept C41008148 @default.
- W4386280904 hasConcept C79974875 @default.
- W4386280904 hasConceptScore W4386280904C103278499 @default.
- W4386280904 hasConceptScore W4386280904C111919701 @default.
- W4386280904 hasConceptScore W4386280904C11413529 @default.
- W4386280904 hasConceptScore W4386280904C115961682 @default.
- W4386280904 hasConceptScore W4386280904C131979681 @default.
- W4386280904 hasConceptScore W4386280904C153180895 @default.
- W4386280904 hasConceptScore W4386280904C154945302 @default.
- W4386280904 hasConceptScore W4386280904C2776135515 @default.
- W4386280904 hasConceptScore W4386280904C2780762811 @default.
- W4386280904 hasConceptScore W4386280904C41008148 @default.
- W4386280904 hasConceptScore W4386280904C79974875 @default.
- W4386280904 hasLocation W43862809041 @default.
- W4386280904 hasOpenAccess W4386280904 @default.
- W4386280904 hasPrimaryLocation W43862809041 @default.
- W4386280904 hasRelatedWork W1892332993 @default.
- W4386280904 hasRelatedWork W2015538044 @default.
- W4386280904 hasRelatedWork W2088785407 @default.
- W4386280904 hasRelatedWork W2383532021 @default.
- W4386280904 hasRelatedWork W2389818373 @default.
- W4386280904 hasRelatedWork W2473504045 @default.
- W4386280904 hasRelatedWork W3004575946 @default.
- W4386280904 hasRelatedWork W3208561927 @default.
- W4386280904 hasRelatedWork W4236152845 @default.
- W4386280904 hasRelatedWork W2155633386 @default.
- W4386280904 isParatext "false" @default.
- W4386280904 isRetracted "false" @default.
- W4386280904 workType "article" @default.