SA3D-L: A lightweight model for 3D object segmentation using neural radiance fields
The Segment Anything Model (SAM) has recently made significant progress in object segmentation within 2D images. However, the task of segmenting objects in a 3D space remains a primary hurdle in computer vision. The neural radiance field (NeRF) utilizes a multilayer perceptron (MLP) to effectively l...
Saved in:
| Published in: | Neurocomputing (Amsterdam) Vol. 623; p. 129420 |
|---|---|
| Main Authors: | , |
| Format: | Journal Article |
| Language: | English |
| Published: |
Elsevier B.V
28.03.2025
|
| Subjects: | |
| ISSN: | 0925-2312 |
| Online Access: | Get full text |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Abstract | The Segment Anything Model (SAM) has recently made significant progress in object segmentation within 2D images. However, the task of segmenting objects in a 3D space remains a primary hurdle in computer vision. The neural radiance field (NeRF) utilizes a multilayer perceptron (MLP) to effectively learn the continuous representation of a 3D scene. Due to its consistent 3D perspectives from various angles, SAM, originally designed for 2D segmentation, can be extended for 3D object segmentation by incorporating NeRF. However, a limitation of NeRF is that the MLP encapsulates the whole scene as a single representation, without distinguishing individual objects. This study introduces a lightweight 3D Segment Anything Model (SA3D-L), which separately represents each segmented object within a scene by modifying the MLP output. Experimental results on established benchmarks revealed that the 3D segmentation representations of objects can be derived from their 2D masks, allowing the independent manipulation of segmented objects and the reconstruction of a new scene. The code is available at: https://github.com/liujian0819/SA3D-L.
•We proposed a method to learn the representation of 3D segmentation from 2D masks.•We utilized a single MLP to represent multiple objects in the scene simultaneously.•We can separately manipulate each segmented object and reconstruct a new scene. |
|---|---|
| AbstractList | The Segment Anything Model (SAM) has recently made significant progress in object segmentation within 2D images. However, the task of segmenting objects in a 3D space remains a primary hurdle in computer vision. The neural radiance field (NeRF) utilizes a multilayer perceptron (MLP) to effectively learn the continuous representation of a 3D scene. Due to its consistent 3D perspectives from various angles, SAM, originally designed for 2D segmentation, can be extended for 3D object segmentation by incorporating NeRF. However, a limitation of NeRF is that the MLP encapsulates the whole scene as a single representation, without distinguishing individual objects. This study introduces a lightweight 3D Segment Anything Model (SA3D-L), which separately represents each segmented object within a scene by modifying the MLP output. Experimental results on established benchmarks revealed that the 3D segmentation representations of objects can be derived from their 2D masks, allowing the independent manipulation of segmented objects and the reconstruction of a new scene. The code is available at: https://github.com/liujian0819/SA3D-L.
•We proposed a method to learn the representation of 3D segmentation from 2D masks.•We utilized a single MLP to represent multiple objects in the scene simultaneously.•We can separately manipulate each segmented object and reconstruct a new scene. |
| ArticleNumber | 129420 |
| Author | Liu, Jian Yu, Zhen |
| Author_xml | – sequence: 1 givenname: Jian orcidid: 0000-0003-2981-2128 surname: Liu fullname: Liu, Jian email: liujian10@zzu.edu.cn organization: School of Computer and Artificial Intelligence, Zhengzhou University, No. 100 Science Avenue, Zhengzhou, Henan, 450001, China – sequence: 2 givenname: Zhen surname: Yu fullname: Yu, Zhen organization: Department of Electrical and Computer Engineering, California State Polytechnic University, Pomona, CA 91768, USA |
| BookMark | eNqFkD1PwzAQhj0UiRb4Bwz-Awn-rJMOSFXLl1SJoTBbjn0pjpIY2SmIf0-qMDHAcjecnrt7nwWa9aEHhK4pySmhy5sm7-FoQ5czwmROWSkYmaE5KZnMGKfsHC1SagihapzN0X6_5ttst8Jr3PrD2_AJp4q74KDFdYiYb3GoGrADTnDooB_M4EOPj8n3BzyeiqbF0Thvegu49tC6dInOatMmuPrpF-j1_u5l85jtnh-eNutdZjlZDpmSxDJHSV0yVSmiyoJX3JWGlo6I2nAlDEjOSlkVxApWOHCVErYQBZdyKSW_QKtpr40hpQi1tn56b4jGt5oSfVKiGz0p0SclelIywuIX_B59Z-LXf9jthMEY7MND1Ml6GLM7H0dJ2gX_94JvhvaAEQ |
| CitedBy_id | crossref_primary_10_1016_j_dsp_2025_105459 crossref_primary_10_1038_s40494_025_02031_z |
| Cites_doi | 10.1109/TPAMI.2021.3098789 10.1109/TPAMI.2016.2572683 10.1109/TPAMI.2017.2699184 10.1109/TPAMI.2016.2577031 |
| ContentType | Journal Article |
| Copyright | 2025 Elsevier B.V. |
| Copyright_xml | – notice: 2025 Elsevier B.V. |
| DBID | AAYXX CITATION |
| DOI | 10.1016/j.neucom.2025.129420 |
| DatabaseName | CrossRef |
| DatabaseTitle | CrossRef |
| DatabaseTitleList | |
| DeliveryMethod | fulltext_linktorsrc |
| Discipline | Computer Science |
| ExternalDocumentID | 10_1016_j_neucom_2025_129420 S092523122500092X |
| GroupedDBID | --- --K --M .DC .~1 0R~ 123 1B1 1~. 1~5 4.4 457 4G. 53G 5VS 7-5 71M 8P~ 9JM 9JN AABNK AACTN AAEDT AAEDW AAIKJ AAKOC AALRI AAOAW AAQFI AAXKI AAXLA AAXUO AAYFN ABBOA ABCQJ ABFNM ABJNI ABMAC ACDAQ ACGFS ACRLP ACZNC ADBBV ADEZE AEBSH AEIPS AEKER AENEX AFJKZ AFTJW AFXIZ AGHFR AGUBO AGWIK AGYEJ AHHHB AHZHX AIALX AIEXJ AIKHN AITUG AKRWK ALMA_UNASSIGNED_HOLDINGS AMRAJ ANKPU AOUOD AXJTR BKOJK BLXMC CS3 DU5 EBS EFJIC EO8 EO9 EP2 EP3 F5P FDB FIRID FNPLU FYGXN G-Q GBLVA GBOLZ IHE J1W KOM M41 MO0 MOBAO N9A O-L O9- OAUVE OZT P-8 P-9 P2P PC. Q38 ROL RPZ SDF SDG SDP SES SEW SPC SPCBC SSN SSV SSZ T5K ZMT ~G- 29N 9DU AAQXK AATTM AAYWO AAYXX ABWVN ABXDB ACLOT ACNNM ACRPL ACVFH ADCNI ADJOM ADMUD ADNMO AEUPX AFPUW AGQPQ AIGII AIIUN AKBMS AKYEP APXCP ASPBG AVWKF AZFZN CITATION EFKBS EFLBG EJD FEDTE FGOYB HLZ HVGLF HZ~ LG9 R2- SBC WUQ XPP ~HD |
| ID | FETCH-LOGICAL-c306t-750c2d10f927b707983b3d9a19d04fa374ae53295b80c428dedb74c8483556553 |
| ISICitedReferencesCount | 2 |
| ISICitedReferencesURI | http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=001403321100001&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D |
| ISSN | 0925-2312 |
| IngestDate | Sat Nov 29 08:19:13 EST 2025 Tue Nov 18 20:40:24 EST 2025 Sat Mar 01 15:45:19 EST 2025 |
| IsPeerReviewed | true |
| IsScholarly | true |
| Keywords | 2D–3D segmentation Multi-layer perceptron (MLP) Neural radiation field (NeRF) Segment Anything Model (SAM) Novel view synthesis |
| Language | English |
| LinkModel | OpenURL |
| MergedId | FETCHMERGED-LOGICAL-c306t-750c2d10f927b707983b3d9a19d04fa374ae53295b80c428dedb74c8483556553 |
| ORCID | 0000-0003-2981-2128 |
| ParticipantIDs | crossref_citationtrail_10_1016_j_neucom_2025_129420 crossref_primary_10_1016_j_neucom_2025_129420 elsevier_sciencedirect_doi_10_1016_j_neucom_2025_129420 |
| PublicationCentury | 2000 |
| PublicationDate | 2025-03-28 |
| PublicationDateYYYYMMDD | 2025-03-28 |
| PublicationDate_xml | – month: 03 year: 2025 text: 2025-03-28 day: 28 |
| PublicationDecade | 2020 |
| PublicationTitle | Neurocomputing (Amsterdam) |
| PublicationYear | 2025 |
| Publisher | Elsevier B.V |
| Publisher_xml | – name: Elsevier B.V |
| References | Yin, Zhou, Zhang, Fang, Xu, Shen, Wang (b33) 2022 Qi, Yi, Su, Guibas (b31) 2017 Kirillov, Mintun, Ravi, Mao, Rolland, Gustafson, Xiao, Whitehead, Berg, Lo (b6) 2023 Ren, Agarwala, Russell, Schwing, Wang (b54) 2022 Ravi, Gabeur, Hu, Hu, Ryali, Ma, Khedr, Rdle, Rolland, Gustafson (b25) 2024 Groueix, Fisher, Kim, Russell, Aubry (b43) 2018 Wu, Ji, Liu, Fu, Xu, Xu, Jin (b59) 2023 Feng, Wang, Wang, Yang, Zheng (b32) 2023 Goel, Sirikonda, Saini, Narayanan (b58) 2022 Feng, Wang, Ma, Yang (b26) 2024 Chen, Zhu, Papandreou, Schroff, Adam (b2) 2018 Salvador, Bellver, Campos, Baradad, Marques, Torres, Giro-I-Nieto (b22) 2017 Peng, Niemeyer, Mescheder, Pollefeys, Geiger (b48) 2020 Sitzmann, Zollhfer, Wetzstein (b51) 2019 Tewari, Thies, Mildenhall, Srinivasan, Tretschk, Wang, Lassner, Sitzmann, Martin-Brualla, Lombardi (b38) 2021 Liu, Zhang, Peng, Shi, Pollefeys, Cui (b49) 2019 Zhu, Zhou, Wang, Hong, Li, Ma, Li, Yang, Lin (b36) 2022; 44 Xiong, Liao, Zhao, Hu, Bai, Yumer, R. (b4) 2019 Rezende, Eslami, Mohamed, Battaglia, Jaderberg, Heess (b40) 2016 Tang, Liu, Zhao, Lin, Lin, Wang, Han (b35) 2020 Vaswani, Shazeer, Parmar, Uszkoreit, Jones, Gomez, Kaiser, Polosukhin (b5) 2017 Wang, Bochkovskiy, Liao (b12) 2022 Zou, Yang, Zhang, Li, Li, Wang, Wang, Gao, Lee (b7) 2023 Chen, Papandreou, Kokkinos, Murphy, Yuille (b19) 2018; 40 Kobayashi, Matsumoto, Sitzmann (b56) 2022 Cai, Vasconcelos (b21) 2017 Bolya, Zhou, Xiao, Lee (b14) 2019; PP Liao, Donné, Geiger (b44) 2018 Berg, Fu, Szegedy, Anguelov, Erhan, Reed, Liu (b13) 2015 Xie, Yao, Sun, Zhou, Zhang (b52) 2020 Gosala, Valada (b37) 2021 Mirzaei, Aumentado-Armstrong, Derpanis, Kelly, Brubaker, Gilitschenski, Levinshtein (b63) 2023 Jiang, Sud, Makadia, Huang, Funkhouser (b47) 2020 Yasir, Ahn (b8) 2024 He, Gkioxari, Dollár, Girshick (b16) 2017 Newell, Huang, Deng (b20) 2016 Wang, Neumann (b28) 2018 Wang, Pan, Cuppens-Boulahia, Cuppens, Roux (b64) 2013 Wu, Allibert, Stolz, Demonceaux (b29) 2021 Cen, Fang, Yang, Xie, Zhang, Shen, Tian (b61) 2023 Hariharan, Arbeláez, Girshick, Malik (b3) 2014 Redmon, Farhadi (b24) 2018 Kerr, Kim, Goldberg, Kanazawa, Tancik (b57) 2023 Qi, Su, Mo, Guibas (b30) 2017 Meng, Wang, Zhou, Shen, Gool, Dai (b34) 2020 Pan, Han, Chen, Tang, Jia (b45) 2020 Zhi, Laidlow, Leutenegger, Davison (b53) 2021 Mildenhall, Srinivasan, Tancik, Barron, Ramamoorthi, Ng (b9) 2020 Tschernezki, Laina, Larlus, Vedaldi (b55) 2022 Shelhamer, Long, Darrell (b11) 2017; 39 Brock, Lim, Ritchie, Weston (b39) 2016 Fang, Yang, Wang, Li, Fang, Shan, Feng, Liu (b23) 2021 Bolya, Zhou, Xiao, Lee (b15) 2020; PP Riegler, Ulusoy, Geiger (b41) 2017 Barron, Mildenhall, Verbin, Srinivasan, Hedman (b62) 2021 Ren, He, Girshick, Sun (b17) 2017; 39 Wang, Zhang, Li, Fu, Liu, Jiang (b46) 2018 Feng, Wang, Quan, Yang (b27) 2024 Wu, Zhang, Xue, Freeman, Tenenbaum (b42) 2016 Gu, Bai, Kong (b1) 2022; 120 Arnab, Torr (b18) 2016 Niemeyer, Mescheder, Oechsle, Geiger (b50) 2020 Dosovitskiy, Beyer, Kolesnikov, Weissenborn, Zhai, Unterthiner, Dehghani, Minderer, Heigold, Gelly (b60) 2021 Cen, Zhou, Fang, Yang, Shen, Xie, Jiang, Zhang, Tian (b10) 2022 Zhu (10.1016/j.neucom.2025.129420_b36) 2022; 44 Tewari (10.1016/j.neucom.2025.129420_b38) 2021 Cen (10.1016/j.neucom.2025.129420_b10) 2022 Wang (10.1016/j.neucom.2025.129420_b12) 2022 Rezende (10.1016/j.neucom.2025.129420_b40) 2016 Wang (10.1016/j.neucom.2025.129420_b28) 2018 Barron (10.1016/j.neucom.2025.129420_b62) 2021 Wang (10.1016/j.neucom.2025.129420_b64) 2013 Tschernezki (10.1016/j.neucom.2025.129420_b55) 2022 Ren (10.1016/j.neucom.2025.129420_b17) 2017; 39 Mirzaei (10.1016/j.neucom.2025.129420_b63) 2023 Peng (10.1016/j.neucom.2025.129420_b48) 2020 Cai (10.1016/j.neucom.2025.129420_b21) 2017 Dosovitskiy (10.1016/j.neucom.2025.129420_b60) 2021 Mildenhall (10.1016/j.neucom.2025.129420_b9) 2020 Newell (10.1016/j.neucom.2025.129420_b20) 2016 Gu (10.1016/j.neucom.2025.129420_b1) 2022; 120 Riegler (10.1016/j.neucom.2025.129420_b41) 2017 Redmon (10.1016/j.neucom.2025.129420_b24) 2018 Shelhamer (10.1016/j.neucom.2025.129420_b11) 2017; 39 Feng (10.1016/j.neucom.2025.129420_b32) 2023 Xie (10.1016/j.neucom.2025.129420_b52) 2020 Groueix (10.1016/j.neucom.2025.129420_b43) 2018 Jiang (10.1016/j.neucom.2025.129420_b47) 2020 Bolya (10.1016/j.neucom.2025.129420_b14) 2019; PP Niemeyer (10.1016/j.neucom.2025.129420_b50) 2020 Feng (10.1016/j.neucom.2025.129420_b27) 2024 Wu (10.1016/j.neucom.2025.129420_b42) 2016 Yasir (10.1016/j.neucom.2025.129420_b8) 2024 Cen (10.1016/j.neucom.2025.129420_b61) 2023 Zou (10.1016/j.neucom.2025.129420_b7) 2023 Hariharan (10.1016/j.neucom.2025.129420_b3) 2014 Bolya (10.1016/j.neucom.2025.129420_b15) 2020; PP Qi (10.1016/j.neucom.2025.129420_b31) 2017 Brock (10.1016/j.neucom.2025.129420_b39) 2016 Berg (10.1016/j.neucom.2025.129420_b13) 2015 Goel (10.1016/j.neucom.2025.129420_b58) 2022 Wu (10.1016/j.neucom.2025.129420_b59) 2023 Fang (10.1016/j.neucom.2025.129420_b23) 2021 Salvador (10.1016/j.neucom.2025.129420_b22) 2017 Arnab (10.1016/j.neucom.2025.129420_b18) 2016 Tang (10.1016/j.neucom.2025.129420_b35) 2020 Feng (10.1016/j.neucom.2025.129420_b26) 2024 He (10.1016/j.neucom.2025.129420_b16) 2017 Meng (10.1016/j.neucom.2025.129420_b34) 2020 Zhi (10.1016/j.neucom.2025.129420_b53) 2021 Gosala (10.1016/j.neucom.2025.129420_b37) 2021 Vaswani (10.1016/j.neucom.2025.129420_b5) 2017 Liao (10.1016/j.neucom.2025.129420_b44) 2018 Pan (10.1016/j.neucom.2025.129420_b45) 2020 Ravi (10.1016/j.neucom.2025.129420_b25) 2024 Liu (10.1016/j.neucom.2025.129420_b49) 2019 Xiong (10.1016/j.neucom.2025.129420_b4) 2019 Chen (10.1016/j.neucom.2025.129420_b19) 2018; 40 Kirillov (10.1016/j.neucom.2025.129420_b6) 2023 Kobayashi (10.1016/j.neucom.2025.129420_b56) 2022 Qi (10.1016/j.neucom.2025.129420_b30) 2017 Chen (10.1016/j.neucom.2025.129420_b2) 2018 Ren (10.1016/j.neucom.2025.129420_b54) 2022 Yin (10.1016/j.neucom.2025.129420_b33) 2022 Kerr (10.1016/j.neucom.2025.129420_b57) 2023 Wang (10.1016/j.neucom.2025.129420_b46) 2018 Wu (10.1016/j.neucom.2025.129420_b29) 2021 Sitzmann (10.1016/j.neucom.2025.129420_b51) 2019 |
| References_xml | – volume: PP year: 2019 ident: b14 article-title: YOLACT: Real-time instance segmentation publication-title: Proceedings of the IEEE International Conference on Computer Vision – year: 2021 ident: b38 article-title: Advances in neural rendering – year: 2018 ident: b28 article-title: Depth-aware CNN for RGB-D segmentation – year: 2022 ident: b33 article-title: ProposalContrast: Unsupervised pre-training for lidar-based 3D object detection publication-title: Proceedings of the European Conference on Computer Vision – volume: 44 start-page: 6807 year: 2022 end-page: 6822 ident: b36 article-title: Cylindrical and asymmetrical 3D convolution networks for LiDAR-based perception publication-title: IEEE Trans. Pattern Anal. Mach. Intell. – volume: PP year: 2020 ident: b15 article-title: YOLACT++: Better real-time instance segmentation publication-title: IEEE Trans. Pattern Anal. Mach. Intell. – year: 2016 ident: b40 article-title: Unsupervised learning of 3D structure from images publication-title: Proceedings of the Conference on Neural Information Processing Systems – year: 2021 ident: b23 article-title: Instances as queries publication-title: Proceedings of the IEEE International Conference on Computer Vision – year: 2019 ident: b49 article-title: DIST: Rendering deep implicit signed distance function with differentiable sphere tracing publication-title: Proceedings of the Conference on Computer Vision and Pattern Recognition – start-page: 297 year: 2014 end-page: 312 ident: b3 article-title: Simultaneous detection and segmentation publication-title: European Conference on Computer Vision – volume: 39 start-page: 1137 year: 2017 end-page: 1149 ident: b17 article-title: Faster R-CNN: Towards real-time object detection with region proposal networks publication-title: IEEE Trans. Pattern Anal. Mach. Intell. – year: 2017 ident: b31 article-title: PointNet++: Deep hierarchical feature learning on point sets in a metric space – year: 2023 ident: b32 article-title: Clustering based point cloud representation learning for 3D analysis publication-title: Proceedings of the International Conference on Computer Vision – year: 2022 ident: b58 article-title: Interactive segmentation of radiance fields – year: 2015 ident: b13 article-title: SSD: Single shot MultiBox detector – year: 2023 ident: b7 article-title: Segment everything everywhere all at once publication-title: Proceedings of the International Conference on Computer Vision and Pattern Recognition – volume: 39 start-page: 640 year: 2017 end-page: 651 ident: b11 article-title: Fully convolutional networks for semantic segmentation publication-title: IEEE Trans. Pattern Anal. Mach. Intell. – year: 2016 ident: b42 article-title: Learning a probabilistic latent space of object shapes via 3D generative-adversarial modeling publication-title: Proceedings of the Conference in Neural Information Processing Systems – year: 2023 ident: b63 article-title: SPIn-NeRF: Multiview segmentation and perceptual inpainting with neural radiance fields. publication-title: Procceding of the Conference on Computer Vision and Pattern Recognition – year: 2023 ident: b57 article-title: Lerf:Language embedded radiance fields – year: 2020 ident: b35 article-title: Searching efficient 3D architectures with sparse point-voxel convolution – year: 2024 ident: b26 article-title: LSK3Dnet: Towards effective and efficient 3D perception with large sparse kernels publication-title: Proceedings of the Conference on Computer Vision and Pattern Recognition – volume: 120 year: 2022 ident: b1 article-title: A review on 2D instance segmentation based on deep neural networks publication-title: Image Vis. Comput. – year: 2019 ident: b51 article-title: Scene representation networks: Continuous 3D-structure-aware neural scene representations publication-title: Proceedings of the Conference in Neural Information Processing Systems – year: 2017 ident: b5 article-title: Attention is all you need – year: 2022 ident: b54 article-title: Neural volumetric object selection publication-title: Procceding of the Conference on Computer Vision and Pattern Recognition – year: 2020 ident: b34 article-title: Weakly Supervised 3D Object Detection from Lidar Point Cloud – year: 2018 ident: b43 article-title: AtlasNet: A papier-mché approach to learning 3D surface generation publication-title: Proceedings of the Conference on Computer Vision and Pattern Recognition – year: 2020 ident: b45 article-title: Deep mesh reconstruction from single RGB images via topology modification networks publication-title: Proceedings of the International Conference on Computer Vision – year: 2024 ident: b25 article-title: SAM 2: Segment anything in images and videos – year: 2022 ident: b55 article-title: Neural feature fusion fields:3d distillation of self-supervised 2d image representations publication-title: International Conference on 3D Vision – year: 2017 ident: b21 article-title: Cascade R-CNN: Delving into high quality object detection – year: 2016 ident: b18 article-title: Bottom-up instance segmentation using deep higher-order CRFs publication-title: British Machine Vision Conference – year: 2017 ident: b30 article-title: PointNet: Deep learning on point sets for 3D classification and segmentation publication-title: Proceedings of the Conference on Computer Vision and Pattern Recognition – year: 2022 ident: b12 article-title: YOLOv7: Trainable bag-of-freebies sets new state-of-the-art for real-time object detectors – year: 2020 ident: b47 article-title: Local implicit grid representations for 3D scenes publication-title: Proceedings of the Conference on Computer Vision and Pattern Recognition – year: 2024 ident: b8 article-title: Deep learning-based 3D instance and semantic segmentation: A review – year: 2021 ident: b62 article-title: Mip-NeRF 360: Unbounded anti-aliased neural radiance fields – year: 2013 ident: b64 article-title: Image quality assessment: From error visibility to structural similarity – start-page: 6620 year: 2017 end-page: 6629 ident: b41 article-title: OctNet: Learning deep 3D representations at high resolutions publication-title: The Conference on Computer Vision and Pattern Recognition – year: 2021 ident: b29 article-title: Depth-adapted CNN for RGB-D cameras publication-title: Proceedings of the Asian Conference on Computer Vision – year: 2018 ident: b44 article-title: Deep marching cubes: Learning explicit surface representations publication-title: Proceedings of the Conference on Computer Vision and Pattern Recognition – year: 2022 ident: b56 article-title: Decomposing nerf for editing via feature field distillation publication-title: The conference on Neural Information Processing Systems – year: 2017 ident: b22 article-title: Recurrent neural networks for semantic instance segmentation – year: 2023 ident: b6 article-title: Segment anything publication-title: Proceedings of the International Conference on Computer Vision – year: 2023 ident: b59 article-title: Medical SAM adapter: Adapting segment anything model for medical image segmentation – start-page: 9404 year: 2019 end-page: 9413 ident: b4 article-title: UPSNet: A unified panoptic segmentation network publication-title: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition – start-page: 801 year: 2018 end-page: 818 ident: b2 article-title: Encoder-decoder with atrous separable convolution for semantic image segmentation publication-title: Proceedings of the European Conference on Computer Vision – year: 2018 ident: b46 article-title: Pixel2Mesh: Generating 3D mesh models from single RGB images publication-title: Proceedings of the European Conference on Computer Vision – year: 2020 ident: b52 article-title: Pix2Vox: Context-aware 3D reconstruction from single and multi-view images publication-title: Proceedings of the International Conference on Computer Vision – volume: 40 start-page: 834 year: 2018 end-page: 848 ident: b19 article-title: DeepLab: Semantic image segmentation with deep convolutional nets, atrous convolution, and fully connected CRFs publication-title: IEEE Trans. Pattern Anal. Mach. Intell. – year: 2018 ident: b24 article-title: YOLOv3: An incremental improvement – year: 2021 ident: b37 article-title: Bird’s-eye-view panoptic segmentation using monocular frontal view images publication-title: IEEE Robot. Autom. Lett. – year: 2017 ident: b16 article-title: Mask R-CNN publication-title: International Conference on Computer Vision – year: 2020 ident: b48 article-title: Convolutional occupancy networks publication-title: Proceedings of the European Conference on Computer Vision – year: 2023 ident: b61 article-title: Segment any 3D Gaussians publication-title: Preceedings of the conference on Neural Information Processing Systems – year: 2021 ident: b53 article-title: In-place scene labelling and understanding with implicit scene representation – year: 2022 ident: b10 article-title: Segment anything in 3d with nerfs publication-title: Preceedings of the conference on Neural Information Processing Systems – year: 2020 ident: b9 article-title: NeRF: Representing scenes as neural radiance fields for view synthesis – year: 2024 ident: b27 article-title: Shape2Scene: 3D scene representation learning through pre-training on shape data – year: 2020 ident: b50 article-title: Differentiable volumetric rendering: Learning implicit 3D representations without 3D supervision publication-title: Proceedings of Conference on Computer Vision and Pattern Recognition – year: 2021 ident: b60 article-title: An image is worth 16x16 words: Transformers for image recognition at scale publication-title: Proceedings of the International Conference on Learning Representations – year: 2016 ident: b20 article-title: Associative embedding: End-to-end learning for joint detection and grouping – year: 2016 ident: b39 article-title: Generative and discriminative voxel modeling with convolutional neural networks publication-title: Comput. Sci. – year: 2017 ident: 10.1016/j.neucom.2025.129420_b31 – year: 2020 ident: 10.1016/j.neucom.2025.129420_b35 – year: 2023 ident: 10.1016/j.neucom.2025.129420_b6 article-title: Segment anything – year: 2020 ident: 10.1016/j.neucom.2025.129420_b45 article-title: Deep mesh reconstruction from single RGB images via topology modification networks – year: 2023 ident: 10.1016/j.neucom.2025.129420_b59 – year: 2021 ident: 10.1016/j.neucom.2025.129420_b62 – year: 2022 ident: 10.1016/j.neucom.2025.129420_b55 article-title: Neural feature fusion fields:3d distillation of self-supervised 2d image representations – year: 2023 ident: 10.1016/j.neucom.2025.129420_b61 article-title: Segment any 3D Gaussians – start-page: 801 year: 2018 ident: 10.1016/j.neucom.2025.129420_b2 article-title: Encoder-decoder with atrous separable convolution for semantic image segmentation – volume: 120 issue: C year: 2022 ident: 10.1016/j.neucom.2025.129420_b1 article-title: A review on 2D instance segmentation based on deep neural networks publication-title: Image Vis. Comput. – year: 2022 ident: 10.1016/j.neucom.2025.129420_b56 article-title: Decomposing nerf for editing via feature field distillation – year: 2018 ident: 10.1016/j.neucom.2025.129420_b43 article-title: AtlasNet: A papier-mché approach to learning 3D surface generation – year: 2017 ident: 10.1016/j.neucom.2025.129420_b21 – volume: 44 start-page: 6807 issue: 10 year: 2022 ident: 10.1016/j.neucom.2025.129420_b36 article-title: Cylindrical and asymmetrical 3D convolution networks for LiDAR-based perception publication-title: IEEE Trans. Pattern Anal. Mach. Intell. doi: 10.1109/TPAMI.2021.3098789 – year: 2017 ident: 10.1016/j.neucom.2025.129420_b16 article-title: Mask R-CNN – volume: PP year: 2019 ident: 10.1016/j.neucom.2025.129420_b14 article-title: YOLACT: Real-time instance segmentation – year: 2024 ident: 10.1016/j.neucom.2025.129420_b27 – year: 2016 ident: 10.1016/j.neucom.2025.129420_b20 – year: 2021 ident: 10.1016/j.neucom.2025.129420_b53 – year: 2017 ident: 10.1016/j.neucom.2025.129420_b30 article-title: PointNet: Deep learning on point sets for 3D classification and segmentation – start-page: 9404 year: 2019 ident: 10.1016/j.neucom.2025.129420_b4 article-title: UPSNet: A unified panoptic segmentation network – year: 2015 ident: 10.1016/j.neucom.2025.129420_b13 – volume: PP issue: 99 year: 2020 ident: 10.1016/j.neucom.2025.129420_b15 article-title: YOLACT++: Better real-time instance segmentation publication-title: IEEE Trans. Pattern Anal. Mach. Intell. – year: 2013 ident: 10.1016/j.neucom.2025.129420_b64 – year: 2019 ident: 10.1016/j.neucom.2025.129420_b49 article-title: DIST: Rendering deep implicit signed distance function with differentiable sphere tracing – year: 2020 ident: 10.1016/j.neucom.2025.129420_b47 article-title: Local implicit grid representations for 3D scenes – year: 2016 ident: 10.1016/j.neucom.2025.129420_b39 article-title: Generative and discriminative voxel modeling with convolutional neural networks publication-title: Comput. Sci. – year: 2024 ident: 10.1016/j.neucom.2025.129420_b25 – year: 2021 ident: 10.1016/j.neucom.2025.129420_b29 article-title: Depth-adapted CNN for RGB-D cameras – year: 2020 ident: 10.1016/j.neucom.2025.129420_b34 – year: 2021 ident: 10.1016/j.neucom.2025.129420_b23 article-title: Instances as queries – year: 2021 ident: 10.1016/j.neucom.2025.129420_b38 – year: 2024 ident: 10.1016/j.neucom.2025.129420_b8 – year: 2018 ident: 10.1016/j.neucom.2025.129420_b44 article-title: Deep marching cubes: Learning explicit surface representations – year: 2022 ident: 10.1016/j.neucom.2025.129420_b54 article-title: Neural volumetric object selection – year: 2023 ident: 10.1016/j.neucom.2025.129420_b32 article-title: Clustering based point cloud representation learning for 3D analysis – year: 2023 ident: 10.1016/j.neucom.2025.129420_b63 article-title: SPIn-NeRF: Multiview segmentation and perceptual inpainting with neural radiance fields. – volume: 39 start-page: 640 issue: 4 year: 2017 ident: 10.1016/j.neucom.2025.129420_b11 article-title: Fully convolutional networks for semantic segmentation publication-title: IEEE Trans. Pattern Anal. Mach. Intell. doi: 10.1109/TPAMI.2016.2572683 – year: 2018 ident: 10.1016/j.neucom.2025.129420_b24 – year: 2018 ident: 10.1016/j.neucom.2025.129420_b46 article-title: Pixel2Mesh: Generating 3D mesh models from single RGB images – year: 2022 ident: 10.1016/j.neucom.2025.129420_b58 – start-page: 6620 year: 2017 ident: 10.1016/j.neucom.2025.129420_b41 article-title: OctNet: Learning deep 3D representations at high resolutions – year: 2020 ident: 10.1016/j.neucom.2025.129420_b50 article-title: Differentiable volumetric rendering: Learning implicit 3D representations without 3D supervision – start-page: 297 year: 2014 ident: 10.1016/j.neucom.2025.129420_b3 article-title: Simultaneous detection and segmentation – year: 2020 ident: 10.1016/j.neucom.2025.129420_b52 article-title: Pix2Vox: Context-aware 3D reconstruction from single and multi-view images – year: 2019 ident: 10.1016/j.neucom.2025.129420_b51 article-title: Scene representation networks: Continuous 3D-structure-aware neural scene representations – year: 2023 ident: 10.1016/j.neucom.2025.129420_b57 – year: 2020 ident: 10.1016/j.neucom.2025.129420_b9 – year: 2022 ident: 10.1016/j.neucom.2025.129420_b33 article-title: ProposalContrast: Unsupervised pre-training for lidar-based 3D object detection – year: 2016 ident: 10.1016/j.neucom.2025.129420_b40 article-title: Unsupervised learning of 3D structure from images – year: 2016 ident: 10.1016/j.neucom.2025.129420_b42 article-title: Learning a probabilistic latent space of object shapes via 3D generative-adversarial modeling – volume: 40 start-page: 834 issue: 4 year: 2018 ident: 10.1016/j.neucom.2025.129420_b19 article-title: DeepLab: Semantic image segmentation with deep convolutional nets, atrous convolution, and fully connected CRFs publication-title: IEEE Trans. Pattern Anal. Mach. Intell. doi: 10.1109/TPAMI.2017.2699184 – year: 2018 ident: 10.1016/j.neucom.2025.129420_b28 – year: 2023 ident: 10.1016/j.neucom.2025.129420_b7 article-title: Segment everything everywhere all at once – year: 2017 ident: 10.1016/j.neucom.2025.129420_b22 – year: 2017 ident: 10.1016/j.neucom.2025.129420_b5 – year: 2016 ident: 10.1016/j.neucom.2025.129420_b18 article-title: Bottom-up instance segmentation using deep higher-order CRFs – year: 2024 ident: 10.1016/j.neucom.2025.129420_b26 article-title: LSK3Dnet: Towards effective and efficient 3D perception with large sparse kernels – year: 2020 ident: 10.1016/j.neucom.2025.129420_b48 article-title: Convolutional occupancy networks – year: 2021 ident: 10.1016/j.neucom.2025.129420_b60 article-title: An image is worth 16x16 words: Transformers for image recognition at scale – year: 2021 ident: 10.1016/j.neucom.2025.129420_b37 article-title: Bird’s-eye-view panoptic segmentation using monocular frontal view images publication-title: IEEE Robot. Autom. Lett. – year: 2022 ident: 10.1016/j.neucom.2025.129420_b12 – year: 2022 ident: 10.1016/j.neucom.2025.129420_b10 article-title: Segment anything in 3d with nerfs – volume: 39 start-page: 1137 issue: 6 year: 2017 ident: 10.1016/j.neucom.2025.129420_b17 article-title: Faster R-CNN: Towards real-time object detection with region proposal networks publication-title: IEEE Trans. Pattern Anal. Mach. Intell. doi: 10.1109/TPAMI.2016.2577031 |
| SSID | ssj0017129 |
| Score | 2.4500718 |
| Snippet | The Segment Anything Model (SAM) has recently made significant progress in object segmentation within 2D images. However, the task of segmenting objects in a... |
| SourceID | crossref elsevier |
| SourceType | Enrichment Source Index Database Publisher |
| StartPage | 129420 |
| SubjectTerms | 2D–3D segmentation Multi-layer perceptron (MLP) Neural radiation field (NeRF) Novel view synthesis Segment Anything Model (SAM) |
| Title | SA3D-L: A lightweight model for 3D object segmentation using neural radiance fields |
| URI | https://dx.doi.org/10.1016/j.neucom.2025.129420 |
| Volume | 623 |
| WOSCitedRecordID | wos001403321100001&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D |
| hasFullText | 1 |
| inHoldings | 1 |
| isFullTextHit | |
| isPrint | |
| journalDatabaseRights | – providerCode: PRVESC databaseName: Elsevier SD Freedom Collection Journals 2021 issn: 0925-2312 databaseCode: AIEXJ dateStart: 19950101 customDbUrl: isFulltext: true dateEnd: 99991231 titleUrlDefault: https://www.sciencedirect.com omitProxy: false ssIdentifier: ssj0017129 providerName: Elsevier |
| link | http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwtV1bb9MwFLbY5YEXBmOIcpMf9lZ5SnzB8d4iOgSompA6UNlL5MQObNrSaW3Hfv6OL8mCioAh7SWqrNhp_Z263zk95zsI7YKDY03CDBGZFoRrVZMMiDNJtRKJi-wrL9f0dSwPD7PpVH2Ovcrnvp2AbJrs-lpd3CvUMAZgu9LZO8DdLQoD8BpAhyvADtd_An6SsxEZh4rzM-d6__TRz9DzxmcVstFwVrr4y3Buv5_H4qNmuPRhAydw6cT-vWYBfOl9itu8z2G9nkflu0HEOEN-7uQWjLOtLq4wPll6C-mZ3zc_cvwjVp_FWAMVLtkq1m77ANhKEUyIJMKNQBN_OVTfhirilQM6xApO9-DDuGwd95A9oBycJrc_SF2a4MQt7Vamvm0Dna6hDSqFgtNrI_94MP3U_V8kUxpUFeNbaYskfSbf6rN-T0J6xOLoMXoUPQKcBySfoAe22UZbbbcNHA_fp2gSgN3HOe7Bij2sGGDFbIQDrLgPK_aw4gArbmHFAdYd9OX9wdG7DyS2xCAV-HYLAvyuoiZNakVl6cQNM1Yyo3SqTMJrzSTXVjCqRJklFXiWxppS8irjQLSBugv2DK03s8Y-R1iXNjWai6Q0nAuuNac1NbWwRlam1HyAWLtHRRX14l3bkrOiTQw8LcLOFm5ni7CzA0S6WRdBL-Uv98t2-4vI-QKXK8Bi_jjzxX_PfIke3hr3K7S-uFza12izulqczC_fRNO6ASL7eaU |
| linkProvider | Elsevier |
| openUrl | ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Ajournal&rft.genre=article&rft.atitle=SA3D-L%3A+A+lightweight+model+for+3D+object+segmentation+using+neural+radiance+fields&rft.jtitle=Neurocomputing+%28Amsterdam%29&rft.au=Liu%2C+Jian&rft.au=Yu%2C+Zhen&rft.date=2025-03-28&rft.pub=Elsevier+B.V&rft.issn=0925-2312&rft.volume=623&rft_id=info:doi/10.1016%2Fj.neucom.2025.129420&rft.externalDocID=S092523122500092X |
| thumbnail_l | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=0925-2312&client=summon |
| thumbnail_m | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=0925-2312&client=summon |
| thumbnail_s | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=0925-2312&client=summon |