Publications
Detailed Information
PeRFception: Perception using Radiance Fields
DC Field | Value | Language |
---|---|---|
dc.contributor.author | Jeong, Yoonwoo | - |
dc.contributor.author | Shin, Seungjoo | - |
dc.contributor.author | Lee, Junha | - |
dc.contributor.author | Choy, Christopher | - |
dc.contributor.author | Anandkumar, Animashree | - |
dc.contributor.author | Cho, Minsu | - |
dc.contributor.author | Park, Jaesik | - |
dc.date.accessioned | 2024-05-09T04:12:04Z | - |
dc.date.available | 2024-05-09T04:12:04Z | - |
dc.date.created | 2024-05-09 | - |
dc.date.issued | 2022 | - |
dc.identifier.citation | Advances in Neural Information Processing Systems, Vol.35 | - |
dc.identifier.issn | 1049-5258 | - |
dc.identifier.uri | https://hdl.handle.net/10371/201284 | - |
dc.description.abstract | The recent progress in implicit 3D representation, i.e., Neural Radiance Fields (NeRFs), has made accurate and photorealistic 3D reconstruction possible in a differentiable manner. This new representation can effectively convey the information of hundreds of high-resolution images in one compact format and allows photorealistic synthesis of novel views. In this work, using the variant of NeRF called Plenoxels, we create the first large-scale radiance fields datasets for perception tasks, called the PeRFception dataset, which consists of two parts that incorporate both object-centric and scene-centric scans for classification and segmentation. It shows a significant memory compression rate (96.4%) from the original dataset, while containing both 2D and 3D information in a unified form. We construct the classification and segmentation models that directly take this radiance fields format as input and also propose a novel augmentation technique to avoid over-fitting on backgrounds of images. The code and data are publicly available in https://postech-cvlab.github.io/PeRFception/. | - |
dc.language | 영어 | - |
dc.publisher | Neural information processing systems foundation | - |
dc.title | PeRFception: Perception using Radiance Fields | - |
dc.type | Article | - |
dc.identifier.doi | 10.48550/arXiv.2208.11537 | - |
dc.citation.journaltitle | Advances in Neural Information Processing Systems | - |
dc.identifier.scopusid | 2-s2.0-85152902964 | - |
dc.citation.volume | 35 | - |
dc.description.isOpenAccess | N | - |
dc.contributor.affiliatedAuthor | Park, Jaesik | - |
dc.type.docType | Conference Paper | - |
dc.description.journalClass | 1 | - |
- Appears in Collections:
- Files in This Item:
- There are no files associated with this item.
Item View & Download Count
Items in S-Space are protected by copyright, with all rights reserved, unless otherwise indicated.