DC Field | Value | Language |
---|---|---|
dc.contributor.author | Choe, Jaesung | - |
dc.contributor.author | Park, Chunghyun | - |
dc.contributor.author | Rameau, Francois | - |
dc.contributor.author | Park, Jaesik | - |
dc.contributor.author | Kweon, In So | - |
dc.date.accessioned | 2023-03-06T00:22:32Z | - |
dc.date.available | 2023-03-06T00:22:32Z | - |
dc.date.created | 2023-03-03 | - |
dc.date.issued | 2022-10-27 | - |
dc.identifier.uri | https://oasis.postech.ac.kr/handle/2014.oak/116832 | - |
dc.description.abstract | MLP-Mixer has newly appeared as a new challenger against the realm of CNNs and Transformer. Despite its simplicity compared to Transformer, the concept of channel-mixing MLPs and token-mixing MLPs achieves noticeable performance in image recognition tasks. Unlike images, point clouds are inherently sparse, unordered and irregular, which limits the direct use of MLP-Mixer for point cloud understanding. To overcome these limitations, we propose PointMixer, a universal point set operator that facilitates information sharing among unstructured 3D point cloud. By simply replacing token-mixing MLPs with Softmax function, PointMixer can “mix” features within/between point sets. By doing so, PointMixer can be broadly used for intra-set, inter-set, and hierarchical-set mixing. We demonstrate that various channel-wise feature aggregation in numerous point sets is better than self-attention layers or dense token-wise interaction in a view of parameter efficiency and accuracy. Extensive experiments show the competitive or superior performance of PointMixer in semantic segmentation, classification, and reconstruction against Transformer-based methods. | - |
dc.language | English | - |
dc.publisher | Springer Science and Business Media Deutschland GmbH | - |
dc.relation.isPartOf | 17th European Conference on Computer Vision, ECCV 2022 | - |
dc.relation.isPartOf | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) | - |
dc.title | PointMixer: MLP-Mixer for Point Cloud Understanding | - |
dc.type | Conference | - |
dc.type.rims | CONF | - |
dc.identifier.bibliographicCitation | 17th European Conference on Computer Vision, ECCV 2022, pp.620 - 640 | - |
dc.identifier.wosid | 000903590200036 | - |
dc.citation.conferenceDate | 2022-10-23 | - |
dc.citation.conferencePlace | IS | - |
dc.citation.endPage | 640 | - |
dc.citation.startPage | 620 | - |
dc.citation.title | 17th European Conference on Computer Vision, ECCV 2022 | - |
dc.contributor.affiliatedAuthor | Park, Jaesik | - |
dc.identifier.scopusid | 2-s2.0-85142704413 | - |
dc.description.journalClass | 1 | - |
dc.description.journalClass | 1 | - |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.
library@postech.ac.kr Tel: 054-279-2548
Copyrights © by 2017 Pohang University of Science ad Technology All right reserved.