File Download
There are no files associated with this item.
Links for fulltext
(May Require Subscription)
- Publisher Website: 10.1109/CVPR46437.2021.00444
- Scopus: eid_2-s2.0-85123182819
- WOS: WOS:000739917304064
- Find via
Supplementary
- Citations:
- Appears in Collections:
Conference Paper: 3D-to-2D Distillation for Indoor Scene Parsing
Title | 3D-to-2D Distillation for Indoor Scene Parsing |
---|---|
Authors | |
Issue Date | 2021 |
Publisher | IEEE Computer Society. The Journal's web site is located at http://ieeexplore.ieee.org/xpl/conhome.jsp?punumber=1000147 |
Citation | Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Nashville, TN, USA, 20-25 June 2021, p. 4464-4474 How to Cite? |
Abstract | Indoor scene semantic parsing from RGB images is very challenging due to occlusions, object distortion, and viewpoint variations. Going beyond prior works that leverage geometry information, typically paired depth maps, we present a new approach, a 3D-to-2D distillation framework, that enables us to leverage 3D features extracted from large-scale 3D data repository (e.g., ScanNet-v2) to enhance 2D features extracted from RGB images. Our work has three novel contributions. First, we distill 3D knowledge from a pretrained 3D network to supervise a 2D network to learn simulated 3D features from 2D features during the training, so the 2D network can infer without requiring 3D data. Second, we design a two-stage dimension normalization scheme to calibrate the 2D and 3D features for better integration. Third, we design a semantic-aware adversarial training model to extend our framework for training with unpaired 3D data. Extensive experiments on various datasets, ScanNet-V2, S3DIS, and NYU-v2, demonstrate the superiority of our approach. Also, experimental results show that our 3D-to-2D distillation improves the model generalization. |
Persistent Identifier | http://hdl.handle.net/10722/306904 |
ISSN | 2023 SCImago Journal Rankings: 10.331 |
ISI Accession Number ID |
DC Field | Value | Language |
---|---|---|
dc.contributor.author | Liu, Z | - |
dc.contributor.author | Qi, X | - |
dc.contributor.author | Fu, CW | - |
dc.date.accessioned | 2021-10-22T07:41:16Z | - |
dc.date.available | 2021-10-22T07:41:16Z | - |
dc.date.issued | 2021 | - |
dc.identifier.citation | Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Nashville, TN, USA, 20-25 June 2021, p. 4464-4474 | - |
dc.identifier.issn | 1063-6919 | - |
dc.identifier.uri | http://hdl.handle.net/10722/306904 | - |
dc.description.abstract | Indoor scene semantic parsing from RGB images is very challenging due to occlusions, object distortion, and viewpoint variations. Going beyond prior works that leverage geometry information, typically paired depth maps, we present a new approach, a 3D-to-2D distillation framework, that enables us to leverage 3D features extracted from large-scale 3D data repository (e.g., ScanNet-v2) to enhance 2D features extracted from RGB images. Our work has three novel contributions. First, we distill 3D knowledge from a pretrained 3D network to supervise a 2D network to learn simulated 3D features from 2D features during the training, so the 2D network can infer without requiring 3D data. Second, we design a two-stage dimension normalization scheme to calibrate the 2D and 3D features for better integration. Third, we design a semantic-aware adversarial training model to extend our framework for training with unpaired 3D data. Extensive experiments on various datasets, ScanNet-V2, S3DIS, and NYU-v2, demonstrate the superiority of our approach. Also, experimental results show that our 3D-to-2D distillation improves the model generalization. | - |
dc.language | eng | - |
dc.publisher | IEEE Computer Society. The Journal's web site is located at http://ieeexplore.ieee.org/xpl/conhome.jsp?punumber=1000147 | - |
dc.relation.ispartof | IEEE Conference on Computer Vision and Pattern Recognition. Proceedings | - |
dc.rights | IEEE Conference on Computer Vision and Pattern Recognition. Proceedings. Copyright © IEEE Computer Society. | - |
dc.rights | ©2021 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works. | - |
dc.title | 3D-to-2D Distillation for Indoor Scene Parsing | - |
dc.type | Conference_Paper | - |
dc.identifier.email | Qi, X: xjqi@eee.hku.hk | - |
dc.identifier.authority | Qi, X=rp02666 | - |
dc.description.nature | link_to_subscribed_fulltext | - |
dc.identifier.doi | 10.1109/CVPR46437.2021.00444 | - |
dc.identifier.scopus | eid_2-s2.0-85123182819 | - |
dc.identifier.hkuros | 328735 | - |
dc.identifier.spage | 4464 | - |
dc.identifier.epage | 4474 | - |
dc.identifier.isi | WOS:000739917304064 | - |
dc.publisher.place | United States | - |