HAC++: Towards 100X Compression of 3D Gaussian Splatting
3D Gaussian Splatting (3DGS) has emerged as a promising representation for novel view synthesis, boosting rapid rendering speed with high fidelity. However, the substantial Gaussians and their associated attributes necessitate effective compression techniques. Nevertheless, the sparse and unorganize...
Saved in:
| Published in: | IEEE transactions on pattern analysis and machine intelligence Vol. 47; no. 11; pp. 10210 - 10226 |
|---|---|
| Main Authors: | , , , , |
| Format: | Journal Article |
| Language: | English |
| Published: |
United States
IEEE
01.11.2025
|
| Subjects: | |
| ISSN: | 0162-8828, 1939-3539, 2160-9292, 1939-3539 |
| Online Access: | Get full text |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Abstract | 3D Gaussian Splatting (3DGS) has emerged as a promising representation for novel view synthesis, boosting rapid rendering speed with high fidelity. However, the substantial Gaussians and their associated attributes necessitate effective compression techniques. Nevertheless, the sparse and unorganized nature of the point cloud of Gaussians (or anchors in our paper) presents challenges for compression. In this paper, we propose HAC++, which explicitly minimizes the representation's entropy during optimization, enabling efficient arithmetic coding after training for compressed storage. Specifically, to reduce entropy, HAC++ leverages the relationships between unorganized anchors and a structured hash grid, utilizing their mutual information for context modeling. Additionally, HAC++ captures intra-anchor contextual relationships to further enhance compression performance. To facilitate entropy coding, we utilize Gaussian distributions to precisely estimate the probability of each quantized attribute, where an adaptive quantization module is proposed to enable high-precision quantization of these attributes for improved fidelity restoration. Moreover, we incorporate an adaptive masking strategy to eliminate non-effective Gaussians and anchors. Overall, HAC++ achieves a remarkable size reduction of over <inline-formula><tex-math notation="LaTeX">100\times</tex-math> <mml:math><mml:mrow><mml:mn>100</mml:mn><mml:mo>×</mml:mo></mml:mrow></mml:math><inline-graphic xlink:href="lin-ieq1-3594066.gif"/> </inline-formula> compared to vanilla 3DGS when averaged on all datasets, while simultaneously improving fidelity. It also delivers more than <inline-formula><tex-math notation="LaTeX">20\times</tex-math> <mml:math><mml:mrow><mml:mn>20</mml:mn><mml:mo>×</mml:mo></mml:mrow></mml:math><inline-graphic xlink:href="lin-ieq2-3594066.gif"/> </inline-formula> size reduction compared to Scaffold-GS. |
|---|---|
| AbstractList | 3D Gaussian Splatting (3DGS) has emerged as a promising representation for novel view synthesis, boosting rapid rendering speed with high fidelity. However, the substantial Gaussians and their associated attributes necessitate effective compression techniques. Nevertheless, the sparse and unorganized nature of the point cloud of Gaussians (or anchors in our paper) presents challenges for compression. In this paper, we propose HAC++, which explicitly minimizes the representation's entropy during optimization, enabling efficient arithmetic coding after training for compressed storage. Specifically, to reduce entropy, HAC++ leverages the relationships between unorganized anchors and a structured hash grid, utilizing their mutual information for context modeling. Additionally, HAC++ captures intra-anchor contextual relationships to further enhance compression performance. To facilitate entropy coding, we utilize Gaussian distributions to precisely estimate the probability of each quantized attribute, where an adaptive quantization module is proposed to enable high-precision quantization of these attributes for improved fidelity restoration. Moreover, we incorporate an adaptive masking strategy to eliminate invalid Gaussians and anchors. Overall, HAC++ achieves a remarkable size reduction of over $100\times$ compared to vanilla 3DGS when averaged on all datasets, while simultaneously improving fidelity. It also delivers more than $20\times$ size reduction compared to Scaffold-GS. Our code is available at https://github.com/YihangChen-ee/HAC-plus.3D Gaussian Splatting (3DGS) has emerged as a promising representation for novel view synthesis, boosting rapid rendering speed with high fidelity. However, the substantial Gaussians and their associated attributes necessitate effective compression techniques. Nevertheless, the sparse and unorganized nature of the point cloud of Gaussians (or anchors in our paper) presents challenges for compression. In this paper, we propose HAC++, which explicitly minimizes the representation's entropy during optimization, enabling efficient arithmetic coding after training for compressed storage. Specifically, to reduce entropy, HAC++ leverages the relationships between unorganized anchors and a structured hash grid, utilizing their mutual information for context modeling. Additionally, HAC++ captures intra-anchor contextual relationships to further enhance compression performance. To facilitate entropy coding, we utilize Gaussian distributions to precisely estimate the probability of each quantized attribute, where an adaptive quantization module is proposed to enable high-precision quantization of these attributes for improved fidelity restoration. Moreover, we incorporate an adaptive masking strategy to eliminate invalid Gaussians and anchors. Overall, HAC++ achieves a remarkable size reduction of over $100\times$ compared to vanilla 3DGS when averaged on all datasets, while simultaneously improving fidelity. It also delivers more than $20\times$ size reduction compared to Scaffold-GS. Our code is available at https://github.com/YihangChen-ee/HAC-plus. 3D Gaussian Splatting (3DGS) has emerged as a promising representation for novel view synthesis, boosting rapid rendering speed with high fidelity. However, the substantial Gaussians and their associated attributes necessitate effective compression techniques. Nevertheless, the sparse and unorganized nature of the point cloud of Gaussians (or anchors in our paper) presents challenges for compression. In this paper, we propose HAC++, which explicitly minimizes the representation's entropy during optimization, enabling efficient arithmetic coding after training for compressed storage. Specifically, to reduce entropy, HAC++ leverages the relationships between unorganized anchors and a structured hash grid, utilizing their mutual information for context modeling. Additionally, HAC++ captures intra-anchor contextual relationships to further enhance compression performance. To facilitate entropy coding, we utilize Gaussian distributions to precisely estimate the probability of each quantized attribute, where an adaptive quantization module is proposed to enable high-precision quantization of these attributes for improved fidelity restoration. Moreover, we incorporate an adaptive masking strategy to eliminate non-effective Gaussians and anchors. Overall, HAC++ achieves a remarkable size reduction of over <inline-formula><tex-math notation="LaTeX">100\times</tex-math> <mml:math><mml:mrow><mml:mn>100</mml:mn><mml:mo>×</mml:mo></mml:mrow></mml:math><inline-graphic xlink:href="lin-ieq1-3594066.gif"/> </inline-formula> compared to vanilla 3DGS when averaged on all datasets, while simultaneously improving fidelity. It also delivers more than <inline-formula><tex-math notation="LaTeX">20\times</tex-math> <mml:math><mml:mrow><mml:mn>20</mml:mn><mml:mo>×</mml:mo></mml:mrow></mml:math><inline-graphic xlink:href="lin-ieq2-3594066.gif"/> </inline-formula> size reduction compared to Scaffold-GS. 3D Gaussian Splatting (3DGS) has emerged as a promising representation for novel view synthesis, boosting rapid rendering speed with high fidelity. However, the substantial Gaussians and their associated attributes necessitate effective compression techniques. Nevertheless, the sparse and unorganized nature of the point cloud of Gaussians (or anchors in our paper) presents challenges for compression. In this paper, we propose HAC++, which explicitly minimizes the representation's entropy during optimization, enabling efficient arithmetic coding after training for compressed storage. Specifically, to reduce entropy, HAC++ leverages the relationships between unorganized anchors and a structured hash grid, utilizing their mutual information for context modeling. Additionally, HAC++ captures intra-anchor contextual relationships to further enhance compression performance. To facilitate entropy coding, we utilize Gaussian distributions to precisely estimate the probability of each quantized attribute, where an adaptive quantization module is proposed to enable high-precision quantization of these attributes for improved fidelity restoration. Moreover, we incorporate an adaptive masking strategy to eliminate non-effective Gaussians and anchors. Overall, HAC++ achieves a remarkable size reduction of over $100\times$100× compared to vanilla 3DGS when averaged on all datasets, while simultaneously improving fidelity. It also delivers more than $20\times$20× size reduction compared to Scaffold-GS. |
| Author | Harandi, Mehrtash Wu, Qianyi Lin, Weiyao Cai, Jianfei Chen, Yihang |
| Author_xml | – sequence: 1 givenname: Yihang orcidid: 0000-0003-1127-1570 surname: Chen fullname: Chen, Yihang email: yhchen.ee@sjtu.edu.cn organization: Shanghai Jiao Tong University, Shanghai, China – sequence: 2 givenname: Qianyi orcidid: 0000-0001-8764-6178 surname: Wu fullname: Wu, Qianyi email: qianyi.wu@monash.edu organization: Monash University, Melbourne, VIC, Australia – sequence: 3 givenname: Weiyao orcidid: 0000-0001-8307-7107 surname: Lin fullname: Lin, Weiyao email: wylin@sjtu.edu.cn organization: Shanghai Jiao Tong University, Shanghai, China – sequence: 4 givenname: Mehrtash orcidid: 0000-0002-6937-6300 surname: Harandi fullname: Harandi, Mehrtash email: mehrtash.harandi@monash.edu organization: Monash University, Melbourne, VIC, Australia – sequence: 5 givenname: Jianfei orcidid: 0000-0002-9444-3763 surname: Cai fullname: Cai, Jianfei email: jianfei.cai@monash.edu organization: Monash University, Melbourne, VIC, Australia |
| BackLink | https://www.ncbi.nlm.nih.gov/pubmed/40742858$$D View this record in MEDLINE/PubMed |
| BookMark | eNpFkF1Lw0AQRRep2A_9AyKSR6Gkzs4mm41vJWpbqChYwbewTWYlki-zCeK_N7VVn4aBc-8wZ8wGZVUSY-ccZpxDeL15mj-sZgjoz4QfeiDlERshl-CGGOKAjYBLdJVCNWRja98BuOeDOGFDDwIPla9GTC3n0XR642yqT92k1uEAr05UFXVD1mZV6VTGEbfOQnf9qkvnuc5122bl2yk7Njq3dHaYE_Zyf7eJlu76cbGK5ms34SG2bgDUn0p5QCFqobQOZJoo7RmJJJWHxoAgCo3xUoWaglQBhiZNhJIaBN-KCbva99ZN9dGRbeMiswnluS6p6mwsUPg9jJ7q0csD2m0LSuO6yQrdfMW_3_YA7oGkqaxtyPwhHOKd0vhHabxTGh-U9qGLfSgjov9Aj0uuAvENv3Ru2g |
| CODEN | ITPIDJ |
| Cites_doi | 10.1109/TVCG.2024.3397828 10.1109/CVPR46437.2021.01453 10.1109/CVPR52729.2023.01981 10.1145/3592433 10.1109/CVPR52729.2023.01201 10.1109/ICME55011.2023.00374 10.1007/s41095-024-0436-y 10.1007/978-3-031-73414-4_25 10.1145/3272127.3275084 10.1109/CVPR52733.2024.01921 10.1109/CVPR52733.2024.02010 10.1002/0471200611 10.1111/cgf.70078 10.1145/3664647.3681468 10.1007/978-3-031-73013-9_2 10.1109/CVPR52688.2022.00539 10.1109/CVPR52733.2024.02052 10.1109/WACV56688.2023.00129 10.1145/214762.214771 10.1109/CVPR.2018.00068 10.1109/ICCV51070.2023.01606 10.1145/3651282 10.1007/978-3-031-72667-5_24 10.1109/CVPR52733.2024.01952 10.1109/CVPR.2016.445 10.1109/CVPR52688.2022.00542 10.1007/978-3-031-19824-3_20 10.1007/978-3-031-73036-8_4 10.1109/CVPR52733.2024.00985 10.1145/3599184.3599188 10.1145/3528223.3530127 10.1145/3072959.3073599 10.1145/3503250 10.1007/978-3-031-72980-5_10 10.1109/CVPR52729.2023.00411 10.1109/CVPR46437.2021.00149 10.1007/978-3-031-73636-0_5 10.1109/TIP.2003.819861 10.1609/aaai.v39i4.32441 10.1007/978-3-031-19824-3_7 10.1109/CVPR42600.2020.00796 10.1109/CVPR52688.2022.00538 |
| ContentType | Journal Article |
| DBID | 97E RIA RIE AAYXX CITATION NPM 7X8 |
| DOI | 10.1109/TPAMI.2025.3594066 |
| DatabaseName | IEEE All-Society Periodicals Package (ASPP) 2005–Present IEEE All-Society Periodicals Package (ASPP) 1998–Present IEEE Electronic Library (IEL) CrossRef PubMed MEDLINE - Academic |
| DatabaseTitle | CrossRef PubMed MEDLINE - Academic |
| DatabaseTitleList | MEDLINE - Academic PubMed |
| Database_xml | – sequence: 1 dbid: NPM name: PubMed url: http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed sourceTypes: Index Database – sequence: 2 dbid: RIE name: IEEE/IET Electronic Library (IEL) (UW System Shared) url: https://ieeexplore.ieee.org/ sourceTypes: Publisher – sequence: 3 dbid: 7X8 name: MEDLINE - Academic url: https://search.proquest.com/medline sourceTypes: Aggregation Database |
| DeliveryMethod | fulltext_linktorsrc |
| Discipline | Engineering Computer Science |
| EISSN | 2160-9292 1939-3539 |
| EndPage | 10226 |
| ExternalDocumentID | 40742858 10_1109_TPAMI_2025_3594066 11106187 |
| Genre | orig-research Journal Article |
| GrantInformation_xml | – fundername: The Australian Research Council Discovery Program through MH grantid: DP230101176 – fundername: National Natural Science Foundation of China grantid: 62325109; U21B2013 funderid: 10.13039/501100001809 |
| GroupedDBID | --- -DZ -~X .DC 0R~ 29I 4.4 53G 5GY 5VS 6IK 97E 9M8 AAJGR AASAJ AAWTH ABAZT ABFSI ABQJQ ABVLG ACGFO ACGFS ACIWK ACNCT ADRHT AENEX AETEA AETIX AGQYO AGSQL AHBIQ AI. AIBXA AKJIK AKQYR ALLEH ALMA_UNASSIGNED_HOLDINGS ASUFR ATWAV BEFXN BFFAM BGNUA BKEBE BPEOZ CS3 DU5 E.L EBS EJD F5P FA8 HZ~ H~9 IBMZZ ICLAB IEDLZ IFIPE IFJZH IPLJI JAVBF LAI M43 MS~ O9- OCL P2P PQQKQ RIA RIE RNI RNS RXW RZB TAE TN5 UHB VH1 XJT ~02 AAYXX CITATION NPM 7X8 |
| ID | FETCH-LOGICAL-c192t-70e742d17e92a38aa76dc8a4f62e6842ff03ee9ff4d82ae7d8029fdc386a031b3 |
| IEDL.DBID | RIE |
| ISSN | 0162-8828 1939-3539 |
| IngestDate | Thu Oct 02 22:11:24 EDT 2025 Tue Oct 07 09:50:08 EDT 2025 Sat Nov 29 07:24:04 EST 2025 Wed Oct 15 14:20:47 EDT 2025 |
| IsPeerReviewed | true |
| IsScholarly | true |
| Issue | 11 |
| Language | English |
| License | https://ieeexplore.ieee.org/Xplorehelp/downloads/license-information/IEEE.html https://doi.org/10.15223/policy-029 https://doi.org/10.15223/policy-037 |
| LinkModel | DirectLink |
| MergedId | FETCHMERGED-LOGICAL-c192t-70e742d17e92a38aa76dc8a4f62e6842ff03ee9ff4d82ae7d8029fdc386a031b3 |
| Notes | ObjectType-Article-1 SourceType-Scholarly Journals-1 ObjectType-Feature-2 content type line 23 |
| ORCID | 0000-0003-1127-1570 0000-0002-6937-6300 0000-0001-8764-6178 0000-0001-8307-7107 0000-0002-9444-3763 |
| PMID | 40742858 |
| PQID | 3235386248 |
| PQPubID | 23479 |
| PageCount | 17 |
| ParticipantIDs | proquest_miscellaneous_3235386248 crossref_primary_10_1109_TPAMI_2025_3594066 pubmed_primary_40742858 ieee_primary_11106187 |
| PublicationCentury | 2000 |
| PublicationDate | 2025-Nov |
| PublicationDateYYYYMMDD | 2025-11-01 |
| PublicationDate_xml | – month: 11 year: 2025 text: 2025-Nov |
| PublicationDecade | 2020 |
| PublicationPlace | United States |
| PublicationPlace_xml | – name: United States |
| PublicationTitle | IEEE transactions on pattern analysis and machine intelligence |
| PublicationTitleAbbrev | TPAMI |
| PublicationTitleAlternate | IEEE Trans Pattern Anal Mach Intell |
| PublicationYear | 2025 |
| Publisher | IEEE |
| Publisher_xml | – name: IEEE |
| References | ref13 ref57 ref56 ref15 ref14 ref53 ref52 ref11 ref10 ref54 ref17 ref16 ref19 Shin (ref27) 2023 Tang (ref30) 2022; 35 ref46 ref45 ref47 Fan (ref18) 2024 Chen (ref49) 2025 Song (ref31) 2024 ref43 Navaneet (ref36) 2024 Paszke (ref51) 2019; 32 ref8 ref7 Ballé (ref48) 2018 ref9 ref4 ref3 ref6 ref5 ref40 Bengio (ref50) 2013 ref35 ref34 ref37 ref33 ref32 ref2 ref1 Yang (ref42) 2024 ref39 Bjontegaard (ref55) 2001 Wang (ref44) 2024 Ali (ref38) 2024 ref24 ref23 ref26 ref25 ref20 ref22 Liu (ref41) 2024 ref21 ref28 ref29 Chen (ref12) 2024 |
| References_xml | – ident: ref13 doi: 10.1109/TVCG.2024.3397828 – ident: ref20 doi: 10.1109/CVPR46437.2021.01453 – ident: ref29 doi: 10.1109/CVPR52729.2023.01981 – ident: ref7 doi: 10.1145/3592433 – ident: ref6 doi: 10.1109/CVPR52729.2023.01201 – ident: ref26 doi: 10.1109/ICME55011.2023.00374 – ident: ref35 doi: 10.1007/s41095-024-0436-y – ident: ref53 doi: 10.1007/978-3-031-73414-4_25 – year: 2024 ident: ref31 article-title: Spc-NeRF: Spatial predictive compression for voxel based radiance field – year: 2001 ident: ref55 article-title: Calculation of average PSNR differences between RD-curves – ident: ref54 doi: 10.1145/3272127.3275084 – ident: ref32 doi: 10.1109/CVPR52733.2024.01921 – ident: ref33 doi: 10.1109/CVPR52733.2024.02010 – year: 2024 ident: ref41 article-title: HEMGS: A hybrid entropy model for 3D Gaussian splatting data compression – ident: ref47 doi: 10.1002/0471200611 – ident: ref14 doi: 10.1111/cgf.70078 – ident: ref45 doi: 10.1145/3664647.3681468 – ident: ref39 doi: 10.1007/978-3-031-73013-9_2 – ident: ref11 doi: 10.1109/CVPR52688.2022.00539 – ident: ref15 doi: 10.1109/CVPR52733.2024.02052 – year: 2013 ident: ref50 article-title: Estimating or propagating gradients through stochastic neurons for conditional computation – ident: ref25 doi: 10.1109/WACV56688.2023.00129 – ident: ref22 doi: 10.1145/214762.214771 – ident: ref57 doi: 10.1109/CVPR.2018.00068 – year: 2024 ident: ref12 article-title: A survey on 3D Gaussian splatting – year: 2024 ident: ref42 article-title: Spectrally pruned Gaussian fields with neural compensation – volume: 32 start-page: 8026 volume-title: Proc. Adv. Neural Inf. Process. Syst. year: 2019 ident: ref51 article-title: PyTorch: An imperative style, high-performance deep learning library – ident: ref28 doi: 10.1109/ICCV51070.2023.01606 – ident: ref37 doi: 10.1145/3651282 – ident: ref23 doi: 10.1007/978-3-031-72667-5_24 – start-page: 55919 volume-title: Proc. Adv. Neural Inf. Process. Syst. year: 2023 ident: ref27 article-title: Binary radiance fields – volume-title: Proc. 13th Int. Conf. Learn. Representations year: 2025 ident: ref49 article-title: Fast feedforward 3D Gaussian splatting compression – ident: ref21 doi: 10.1109/CVPR52733.2024.01952 – start-page: 51532 volume-title: Proc. Adv. neural Inf. Process. Syst. year: 2024 ident: ref44 article-title: ContextGS: Compact 3D Gaussian splatting with anchor level context model – ident: ref8 doi: 10.1109/CVPR.2016.445 – start-page: 330 volume-title: Proc. Eur. Conf. Comput. Vis. year: 2024 ident: ref36 article-title: Compact3D: Compressing Gaussian splat radiance field models with vector quantization – volume-title: Proc. Int. Conf. Learn. Representations year: 2018 ident: ref48 article-title: Variational image compression with a scale hyperprior – ident: ref2 doi: 10.1109/CVPR52688.2022.00542 – ident: ref5 doi: 10.1007/978-3-031-19824-3_20 – ident: ref17 doi: 10.1007/978-3-031-73036-8_4 – volume-title: Proc. BMVC year: 2024 ident: ref38 article-title: Trimming the fat: Efficient compression of 3D Gaussian splats through pruning – volume: 35 start-page: 14798 year: 2022 ident: ref30 article-title: Compressible-composable NeRF via rank-residual decomposition publication-title: Adv. Neural Inf. Process. Syst. – ident: ref16 doi: 10.1109/CVPR52733.2024.00985 – ident: ref46 doi: 10.1145/3599184.3599188 – ident: ref4 doi: 10.1145/3528223.3530127 – ident: ref52 doi: 10.1145/3072959.3073599 – ident: ref1 doi: 10.1145/3503250 – ident: ref43 doi: 10.1007/978-3-031-72980-5_10 – ident: ref24 doi: 10.1109/CVPR52729.2023.00411 – ident: ref9 doi: 10.1109/CVPR46437.2021.00149 – ident: ref40 doi: 10.1007/978-3-031-73636-0_5 – ident: ref56 doi: 10.1109/TIP.2003.819861 – start-page: 140138 volume-title: Proc. Adv. Neural Inf. Process. Syst. year: 2024 ident: ref18 article-title: LightGaussian: Unbounded 3D Gaussian compression with 15x reduction and 200 FPS – ident: ref34 doi: 10.1609/aaai.v39i4.32441 – ident: ref10 doi: 10.1007/978-3-031-19824-3_7 – ident: ref19 doi: 10.1109/CVPR42600.2020.00796 – ident: ref3 doi: 10.1109/CVPR52688.2022.00538 |
| SSID | ssj0014503 |
| Score | 2.5037978 |
| Snippet | 3D Gaussian Splatting (3DGS) has emerged as a promising representation for novel view synthesis, boosting rapid rendering speed with high fidelity. However,... |
| SourceID | proquest pubmed crossref ieee |
| SourceType | Aggregation Database Index Database Publisher |
| StartPage | 10210 |
| SubjectTerms | 3D gaussian splatting (3DGS) Adaptation models compression context model Context modeling Entropy Entropy coding Mutual information Neural radiance field Redundancy Rendering (computer graphics) Three-dimensional displays Training |
| Title | HAC++: Towards 100X Compression of 3D Gaussian Splatting |
| URI | https://ieeexplore.ieee.org/document/11106187 https://www.ncbi.nlm.nih.gov/pubmed/40742858 https://www.proquest.com/docview/3235386248 |
| Volume | 47 |
| hasFullText | 1 |
| inHoldings | 1 |
| isFullTextHit | |
| isPrint | |
| journalDatabaseRights | – providerCode: PRVIEE databaseName: IEEE/IET Electronic Library (IEL) (UW System Shared) customDbUrl: eissn: 2160-9292 dateEnd: 99991231 omitProxy: false ssIdentifier: ssj0014503 issn: 0162-8828 databaseCode: RIE dateStart: 19790101 isFulltext: true titleUrlDefault: https://ieeexplore.ieee.org/ providerName: IEEE |
| link | http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV1LT8MwDLYAcYADzwHjMQWJ21TokrRJuE28hgRoEkParcpSV0JCHWIbvx8nbYELB249NE3rz6nt2M4HcEaRv0M3UZFz2kVSo4yMophHW-dcnGvSg3Bk_oN6etLjsRnWzeqhFwYRQ_EZnvvLkMvPp27ht8ouaF2S-dFqGZaVSqtmre-UgUwCDTK5MDQTxRFNh0xsLkbD_uM9xYI8OReJIRPmiYukjwq1p3r_ZZACw8rfzmYwOreb_3zdLdiovUvWr9RhG5aw3IHNhrmB1Qt5B9Z_HUO4C3rQv-p2L9kolNDOWC-Ox8wPqmpkSzYtmLhmd3Yx8y2X7Pn9zYZy6Ra83N6MrgZRzagQOfLk5pGKkT467yk03AptrUpzp60sUo4-IVcUsUA0RSFzzS0qwoqbIndCp5ZW_0TswUo5LfEAmEx6SqTIJ57lJu05kxirnCrchGttULah24g1e68OzshCwBGbLOCReTyyGo82tLz8fu6sRdeG0waKjPTeJzNsidPFLBNc0L865VK3Yb_C6Ht0A-3hH089gjU_edVSeAwr848FnsCq-5y_zj46pFxj3QnK9QUdT8Wf |
| linkProvider | IEEE |
| linkToHtml | http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV1LT8MwDLZ4ScCB54DxDBK3qVuXpE3CbeI1xDZNYki7VVnqSkioQ2zj95OkLXDhwK2Hpk39ObUd2_kArmzkb9BMRGCMNAGXyAMlbMwjtTEmTKXVA39kfk8MBnI8VsOyWd33wiCiLz7Dprv0ufx0ahZuq6xl16U1P1Isw2rEOQ2Ldq3vpAGPPBGydWLsu2wkUfXIhKo1Gnb6jzYapFGTRcoaMUddxF1cKB3Z-y-T5DlW_nY3vdm53_7nhHdgq_QvSadQiF1YwnwPtivuBlIu5T3Y_HUQ4T7Ibuem0bgmI19EOyPtMBwTN6ioks3JNCPsljzoxcw1XZLn9zftC6Zr8HJ_N7rpBiWnQmCsLzcPRIj2o9O2QEU1k1qLODVS8yym6FJyWRYyRJVlPJVUo7BoUZWlhslY2_U_YQewkk9zPALCo7ZgMdKJ47mJ20ZFSgsjMjOhUirkdWhUYk3ei6MzEh9yhCrxeCQOj6TEow41J7-fO0vR1eGygiKxmu_SGTrH6WKWMMrs3zqmXNbhsMDoe3QF7fEfT72A9e6o30t6j4OnE9hwEykaDE9hZf6xwDNYM5_z19nHuVexL-zIx_4 |
| 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=HAC%2B%2B%3A+Towards+100X+Compression+of+3D+Gaussian+Splatting&rft.jtitle=IEEE+transactions+on+pattern+analysis+and+machine+intelligence&rft.au=Chen%2C+Yihang&rft.au=Wu%2C+Qianyi&rft.au=Lin%2C+Weiyao&rft.au=Harandi%2C+Mehrtash&rft.date=2025-11-01&rft.pub=IEEE&rft.issn=0162-8828&rft.volume=47&rft.issue=11&rft.spage=10210&rft.epage=10226&rft_id=info:doi/10.1109%2FTPAMI.2025.3594066&rft_id=info%3Apmid%2F40742858&rft.externalDocID=11106187 |
| thumbnail_l | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=0162-8828&client=summon |
| thumbnail_m | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=0162-8828&client=summon |
| thumbnail_s | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=0162-8828&client=summon |