SpecTrHuMS: Spectral transformer for human mesh sequence learning
We present SpecTrHuMS, a Spectral Transformer for 3D triangular Human Mesh Sequence learning which combines known deep learning models with spectral mesh processing to capture characteristics of 3D shapes as well as temporal dependencies between the frames. Unlike previous works in this field, our a...
Gespeichert in:
| Veröffentlicht in: | Computers & graphics Jg. 115; S. 191 - 203 |
|---|---|
| Hauptverfasser: | , , , |
| Format: | Journal Article |
| Sprache: | Englisch |
| Veröffentlicht: |
Elsevier Ltd
01.10.2023
Elsevier |
| Schlagworte: | |
| ISSN: | 0097-8493 |
| Online-Zugang: | Volltext |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| Abstract | We present SpecTrHuMS, a Spectral Transformer for 3D triangular Human Mesh Sequence learning which combines known deep learning models with spectral mesh processing to capture characteristics of 3D shapes as well as temporal dependencies between the frames. Unlike previous works in this field, our approach is able to work directly with a compressed representation of the geometry, the spectral coefficients, rather than relying solely on skeleton joints that does not contain surface information. The vertices of each mesh of a sequence are first projected on the eigenvectors of the Graph Laplacian computed from the common triangulation. A convolutional encoder then encodes each frame into lower dimensional latent variables that preserve as much as possible the spectral information. These latent variables are next passed through a transformer architecture so that the model understands the context of the sequence and learns temporal dependencies between the frames. Each frame of the transformer’s output is then decoded by a convolutional decoder which aims to reconstruct the input spectral coefficients. Finally, all frames are transformed back into the spatial domain, resulting in a general process able to treat 4D surfaces with a constant connectivity. Our method is evaluated on a prediction task on AMASS, a dataset of human surface sequences, showing the ability of our model to produce realistic movements while preserving the identity of a subject, and showing that this work is a significant step towards efficient and high-quality representation of triangular mesh sequences with constant connectivity. Additional experiments show that our model can be easily extended to other tasks such as long term prediction, completion and that it is generalizable to other datasets with constant connectivity. This work opens up new possibilities for applications in the fields of animation, virtual reality, and computer graphics. Pretrained models, the code to train them and the code to create datasets will be made publicly available.
[Display omitted]
•This work focuses on generating human triangular mesh sequences.•It consists in the association of a convolutional autoencoder and a transformer.•Spectral coefficients computed from the Graph Laplacian are given as input.•The context of the sequence is understood using a transformer.•The trained model is able to predict movements and preserve the identity of subjects. |
|---|---|
| AbstractList | We present SpecTrHuMS, a Spectral Transformer for 3D triangular Human Mesh Sequence learning which combines known deep learning models with spectral mesh processing to capture characteristics of 3D shapes as well as temporal dependencies between the frames. Unlike previous works in this field, our approach is able to work directly with a compressed representation of the geometry, the spectral coefficients, rather than relying solely on skeleton joints that does not contain surface information. The vertices of each mesh of a sequence are first projected on the eigenvectors of the Graph Laplacian computed from the common triangulation. A convolutional encoder then encodes each frame into lower dimensional latent variables that preserve as much as possible the spectral information. These latent variables are next passed through a transformer architecture so that the model understands the context of the sequence and learns temporal dependencies between the frames. Each frame of the transformer’s output is then decoded by a convolutional decoder which aims to reconstruct the input spectral coefficients. Finally, all frames are transformed back into the spatial domain, resulting in a general process able to treat 4D surfaces with a constant connectivity. Our method is evaluated on a prediction task on AMASS, a dataset of human surface sequences, showing the ability of our model to produce realistic movements while preserving the identity of a subject, and showing that this work is a significant step towards efficient and high-quality representation of triangular mesh sequences with constant connectivity. Additional experiments show that our model can be easily extended to other tasks such as long term prediction, completion and that it is generalizable to other datasets with constant connectivity. This work opens up new possibilities for applications in the fields of animation, virtual reality, and computer graphics. Pretrained models, the code to train them and the code to create datasets will be made publicly available. We present SpecTrHuMS, a Spectral Transformer for 3D triangular Human Mesh Sequence learning which combines known deep learning models with spectral mesh processing to capture characteristics of 3D shapes as well as temporal dependencies between the frames. Unlike previous works in this field, our approach is able to work directly with a compressed representation of the geometry, the spectral coefficients, rather than relying solely on skeleton joints that does not contain surface information. The vertices of each mesh of a sequence are first projected on the eigenvectors of the Graph Laplacian computed from the common triangulation. A convolutional encoder then encodes each frame into lower dimensional latent variables that preserve as much as possible the spectral information. These latent variables are next passed through a transformer architecture so that the model understands the context of the sequence and learns temporal dependencies between the frames. Each frame of the transformer’s output is then decoded by a convolutional decoder which aims to reconstruct the input spectral coefficients. Finally, all frames are transformed back into the spatial domain, resulting in a general process able to treat 4D surfaces with a constant connectivity. Our method is evaluated on a prediction task on AMASS, a dataset of human surface sequences, showing the ability of our model to produce realistic movements while preserving the identity of a subject, and showing that this work is a significant step towards efficient and high-quality representation of triangular mesh sequences with constant connectivity. Additional experiments show that our model can be easily extended to other tasks such as long term prediction, completion and that it is generalizable to other datasets with constant connectivity. This work opens up new possibilities for applications in the fields of animation, virtual reality, and computer graphics. Pretrained models, the code to train them and the code to create datasets will be made publicly available. [Display omitted] •This work focuses on generating human triangular mesh sequences.•It consists in the association of a convolutional autoencoder and a transformer.•Spectral coefficients computed from the Graph Laplacian are given as input.•The context of the sequence is understood using a transformer.•The trained model is able to predict movements and preserve the identity of subjects. |
| Author | Denis, Florence Dupont, Florent Lavoué, Guillaume Lemeunier, Clément |
| Author_xml | – sequence: 1 givenname: Clément orcidid: 0000-0002-2853-6267 surname: Lemeunier fullname: Lemeunier, Clément email: clement.lemeunier@liris.cnrs.fr organization: Univ Lyon, CNRS, INSA Lyon, UCBL, LIRIS, UMR5205, F-69622 Villeurbanne, France – sequence: 2 givenname: Florence surname: Denis fullname: Denis, Florence email: florence.denis@liris.cnrs.fr organization: Univ Lyon, UCBL, CNRS, INSA Lyon, LIRIS, UMR5205, F-69622 Villeurbanne, France – sequence: 3 givenname: Guillaume surname: Lavoué fullname: Lavoué, Guillaume email: guillaume.lavoue@liris.cnrs.fr organization: Univ Lyon, Centrale Lyon, CNRS, INSA Lyon, UCBL, LIRIS, UMR5205, ENISE, F-42023 Saint-Étienne, France – sequence: 4 givenname: Florent surname: Dupont fullname: Dupont, Florent email: florent.dupont@liris.cnrs.fr organization: Univ Lyon, UCBL, CNRS, INSA Lyon, LIRIS, UMR5205, F-69622 Villeurbanne, France |
| BackLink | https://hal.science/hal-04154623$$DView record in HAL |
| BookMark | eNp9kD1PwzAQhj0UibbwA9iyMiScP_JRmKoKKFIRQ8tsOc6ldZU4xU4r8e9xVMTA0OXe0-mek-6ZkJHtLBJyRyGhQLOHfaLVNmHAeAJ5AkBHZAwwy-NCzPg1mXi_BwDGMjEm8_UB9cYtj-_rx2joe6eaKBTr68616KIQ0e7YKhu16HeRx68jWo1Rg8pZY7c35KpWjcfb35ySz5fnzWIZrz5e3xbzVax5Lvq41GVaKNRlXmBdqpRXUHGezWiRsqqCtEzrFGuNSlNai4KmRc5FSRVnilYgGJ-S-_PdnWrkwZlWuW_ZKSOX85UcZiBoKjLGTzTs0vOudp33Dus_gIIcHMm9DI7k4EhCLoOjwOT_GG161ZvOBhumuUg-nUkM758MOum1GRxVxgWhsurMBfoHHmiEzA |
| CitedBy_id | crossref_primary_10_3390_electronics13040720 crossref_primary_10_1016_j_cag_2023_12_007 crossref_primary_10_1016_j_cag_2023_10_018 crossref_primary_10_3390_app14072821 |
| Cites_doi | 10.1109/ICCV.2019.00554 10.1080/00401706.1962.10490022 10.1111/cgf.14012 10.1109/CVPR.2018.00275 10.1109/TSP.2018.2879624 10.1007/s11263-021-01492-6 10.1109/MSP.2017.2693418 10.1109/CVPR.2019.01123 10.1007/978-3-030-01219-9_43 10.1109/TPAMI.2013.248 10.1007/978-3-030-11015-4_26 10.1111/cgf.14599 10.1109/TPAMI.2020.3005434 10.1145/2908736 10.1145/3306346.3322959 10.1007/s11263-021-01483-7 10.1145/2816795.2818013 10.1016/j.cag.2016.05.018 10.1016/j.cag.2022.07.011 10.1145/3507905 10.1145/2185520.2185526 10.24963/ijcai.2018/130 10.1007/978-3-031-20047-2_28 |
| ContentType | Journal Article |
| Copyright | 2023 Elsevier Ltd Distributed under a Creative Commons Attribution 4.0 International License |
| Copyright_xml | – notice: 2023 Elsevier Ltd – notice: Distributed under a Creative Commons Attribution 4.0 International License |
| DBID | AAYXX CITATION 1XC VOOES |
| DOI | 10.1016/j.cag.2023.07.001 |
| DatabaseName | CrossRef Hyper Article en Ligne (HAL) Hyper Article en Ligne (HAL) (Open Access) |
| DatabaseTitle | CrossRef |
| DatabaseTitleList | |
| DeliveryMethod | fulltext_linktorsrc |
| Discipline | Engineering Computer Science |
| EndPage | 203 |
| ExternalDocumentID | oai:HAL:hal-04154623v1 10_1016_j_cag_2023_07_001 S0097849323001309 |
| GroupedDBID | --K --M -~X .DC .~1 0R~ 1B1 1~. 1~5 29F 4.4 457 4G. 5GY 5VS 6TJ 7-5 71M 8P~ 9JN AAEDT AAEDW AAIKJ AAKOC AALRI AAOAW AAQFI AAQXK AATTM AAXKI AAXUO AAYFN AAYWO ABAOU ABBOA ABDPE ABEFU ABJNI ABMAC ABWVN ABXDB ACDAQ ACGFS ACLOT ACNNM ACRLP ACRPL ACZNC ADBBV ADEZE ADGUI ADJOM ADMUD ADNMO AEBSH AEIPS AEKER AFFNX AFJKZ AFTJW AGHFR AGQPQ AGSOS AGUBO AGYEJ AHHHB AHZHX AI. AIALX AIEXJ AIGVJ AIIUN AIKHN AITUG ALMA_UNASSIGNED_HOLDINGS AMRAJ ANKPU AOUOD APXCP ARUGR ASPBG AVWKF AXJTR AZFZN BKOJK BLXMC CS3 EBS EFJIC EFKBS EFLBG EJD EO8 EO9 EP2 EP3 F5P FDB FEDTE FGOYB FIRID FNPLU FYGXN G-Q GBLVA GBOLZ HLZ HVGLF HZ~ H~9 IHE J1W K-O KOM LG9 M41 MHUIS MO0 N9A O-L O9- OAUVE OZT P-8 P-9 PC. Q38 R2- ROL RPZ SBC SDF SDG SDP SES SEW SPC SPCBC SSV SSW SSZ T5K TN5 UHS VH1 VOH WH7 WUQ XPP ZMT ZY4 ~02 ~G- ~HD 9DU AAYXX CITATION 1XC VOOES |
| ID | FETCH-LOGICAL-c374t-bcb58aecb78efba53d0d33691852dd05b5f5efceac11f48158734b1a32a1d0423 |
| ISICitedReferencesCount | 5 |
| ISICitedReferencesURI | http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=001056807800001&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D |
| ISSN | 0097-8493 |
| IngestDate | Tue Oct 14 21:03:23 EDT 2025 Tue Nov 18 22:19:37 EST 2025 Sat Nov 29 07:24:38 EST 2025 Sat Oct 04 17:01:45 EDT 2025 |
| IsDoiOpenAccess | true |
| IsOpenAccess | true |
| IsPeerReviewed | true |
| IsScholarly | true |
| Keywords | Transformers Geometric deep learning Autoencoder Human body triangular mesh sequences Spectral analysis Geometric Deep Learning |
| Language | English |
| License | Distributed under a Creative Commons Attribution 4.0 International License: http://creativecommons.org/licenses/by/4.0 |
| LinkModel | OpenURL |
| MergedId | FETCHMERGED-LOGICAL-c374t-bcb58aecb78efba53d0d33691852dd05b5f5efceac11f48158734b1a32a1d0423 |
| ORCID | 0000-0002-2853-6267 0000-0003-3988-6702 0000-0001-6611-4420 |
| OpenAccessLink | https://hal.science/hal-04154623 |
| PageCount | 13 |
| ParticipantIDs | hal_primary_oai_HAL_hal_04154623v1 crossref_primary_10_1016_j_cag_2023_07_001 crossref_citationtrail_10_1016_j_cag_2023_07_001 elsevier_sciencedirect_doi_10_1016_j_cag_2023_07_001 |
| PublicationCentury | 2000 |
| PublicationDate | 2023-10-01 |
| PublicationDateYYYYMMDD | 2023-10-01 |
| PublicationDate_xml | – month: 10 year: 2023 text: 2023-10-01 day: 01 |
| PublicationDecade | 2020 |
| PublicationTitle | Computers & graphics |
| PublicationYear | 2023 |
| Publisher | Elsevier Ltd Elsevier |
| Publisher_xml | – name: Elsevier Ltd – name: Elsevier |
| References | Sofianos, Sampieri, Franco, Galasso (b7) 2021 Levie, Monti, Bresson, Bronstein (b33) 2019; 67 Sharp, Attaiki, Crane, Ovsjanikov (b38) 2022; 41 Gao, Lai, Liang, Chen, Xia (b39) 2016; 35 Martinez, Black, Romero (b50) 2017 Mao, Liu, Salzmann, Li (b5) 2021; 129 Lebailly, Kiciroglu, Salzmann, Fua, Wang (b57) 2021 Lim, Dielen, Campen, Kobbelt (b23) 2019 Guo, Du, Shen, Lepetit, Alameda-Pineda, Moreno-Noguer (b6) 2023 Devlin, Chang, Lee, Toutanova (b15) 2018 Lehrmann, Gehler, Nowozin (b47) 2014 Fey, Lenssen, Weichert, Muller (b22) 2018 Mahmood N, Ghorbani N, Troje NF, Pons-Moll G, Black MJ. AMASS: Archive of Motion Capture as Surface Shapes. In: International conference on computer vision. 2019, p. 5442–51. Guo, Wang, Hu, Liu, Liu, Bennamoun (b16) 2021; 43 Jain, Zamir, Savarese, Saxena (b49) 2016 Mao, Liu, Salzmann (b4) 2020 Eisenberger, Novotny, Kerchenbaum, Labatut, Neverova, Cremers (b43) 2021 Qi CR, Su H, Mo K, Guibas LJ. PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation. In: Proceedings of the IEEE conference on computer vision and pattern recognition. CVPR, 2017. Cai, Huang, Wang, Cham, Cai, Yuan (b62) 2020 Bouritsas, Bokhnyak, Ploumpis, Zafeiriou, Bronstein (b24) 2019 Mao, Liu, Salzmann, Li (b3) 2019 Lemeunier, Denis, Lavoué, Dupont (b14) 2022; 107 Masci, Boscaini, Bronstein, Vandergheynst (b20) 2015 Sun, Rooke, Charton, He, Lu, Baek (b34) 2020; 39 Defferrard, Bresson, Vandergheynst (b30) 2016 Kipf, Welling (b31) 2017 Butepage, Black, Kragic, Kjellstrom (b53) 2017 Hernandez, Gall, Moreno (b55) 2019 Verma N, Boyer E, Verbeek J. FeaStNet: Feature-Steered Graph Convolutions for 3D Shape Analysis. In: CVPR - IEEE conference on computer vision & pattern recognition. Salt Lake City, United States; 2018, URL. Ranjan, Bolkart, Sanyal, Black (b32) 2018 Petrovich, Black, Varol (b64) 2022 Bronstein, Bruna, LeCun, Szlam, Vandergheynst (b1) 2017; 34 Ma, Nie, Long, Zhang, Li (b58) 2022 Welford (b71) 1962; 4 Bruna, Zaremba, Szlam, Lecun (b29) 2014 Gao, Lai, Yang, Zhang, Kobbelt, Xia (b41) 2017 Abrevaya, Manandhar, Hétroy-Wheeler, Wuhrer (b68) 2016; 58 Aksan, Kaufmann, Cao, Hilliges (b61) 2021 Pavlakos G, Choutas V, Ghorbani N, Bolkart T, Osman AAA, Tzionas D, et al. Expressive Body Capture: 3D Hands, Face, and Body from a Single Image. In: Proceedings IEEE conf. on computer vision and pattern recognition. CVPR, 2019, p. 10975–85. Liu, Wu, Jin, Liu, Lu, Zimmermann (b52) 2019 Rakotosaona, Ovsjanikov (b12) 2020 Hanocka, Hertz, Fish, Giryes, Fleishman, Cohen-Or (b25) 2019; 38 Hartman, Pierson, Bauer, Charon, Daoudi (b44) 2022 Marsot, Wuhrer, Franco, Durocher (b9) 2021 Marin, Rampini, Castellani, Rodola, Ovsjanikov, Melzi (b35) 2020 Aumentado-Armstrong, Tsogkas, Jepson, Dickinson (b19) 2019 Cosmo, Norelli, Halimi, Kimmel, Rodolà (b11) 2020 Fragkiadaki, Levine, Felsen, Malik (b48) 2015 Yuan, Lai, Yang, Fu, Gao (b42) 2019 Tang, Ma, Liu, Zheng (b60) 2018 Ovsjanikov, Ben-Chen, Solomon, Butscher, Guibas (b73) 2012; 31 Ionescu, Papava, Olaru, Sminchisescu (b70) 2014; 36 Gong, Chen, Bronstein, Zafeiriou (b13) 2019 Tan, Gao, Lai, Xia (b40) 2017 Li, Zhang, Lee, Lee (b2) 2018 Duan, Shi, Zou, Lin, Qian, Zhang (b63) 2021 Loper, Mahmood, Romero, Pons-Moll, Black (b10) 2015; 34 Vaswani, Shazeer, Parmar, Uszkoreit, Jones, Gomez (b59) 2017 Community (b72) 2018 Sohn, Lee, Yan (b66) 2015 Gui, Wang, Liang, Moura (b56) 2018 Tevet, Raab, Gordon, Shafir, Cohen-Or, Bermano (b65) 2022 Zuffi, Kanazawa, Jacobs, Black (b67) 2017 Marin, Rampini, Castellani, Rodolà, Ovsjanikov, Melzi (b36) 2021; 129 Taylor, Hinton, Roweis (b46) 2006 Butepage, Kjellstrom, Kragic (b54) 2018 Pegoraro, Melzi, Castellani, Marin, Rodolà (b37) 2022; 41 Chiu, Adeli, Wang, Huang, Niebles (b51) 2019 Zhong, Hu, Zhang, Ye, Xia (b8) 2022 Milano, Loquercio, Rosinol, Scaramuzza, Carlone (b28) 2020 Boscaini, Masci, Rodoià, Bronstein (b21) 2016 Qi, Yi, Su, Guibas (b18) 2017 Huang, Huang, Sun, Zhang, Jiang, Bajaj (b26) 2021 Wang, Hertzmann, Fleet (b45) 2005 Cosmo (10.1016/j.cag.2023.07.001_b11) 2020 Yuan (10.1016/j.cag.2023.07.001_b42) 2019 Liu (10.1016/j.cag.2023.07.001_b52) 2019 Martinez (10.1016/j.cag.2023.07.001_b50) 2017 Petrovich (10.1016/j.cag.2023.07.001_b64) 2022 Lim (10.1016/j.cag.2023.07.001_b23) 2019 Sharp (10.1016/j.cag.2023.07.001_b38) 2022; 41 Ionescu (10.1016/j.cag.2023.07.001_b70) 2014; 36 Duan (10.1016/j.cag.2023.07.001_b63) 2021 Milano (10.1016/j.cag.2023.07.001_b28) 2020 Devlin (10.1016/j.cag.2023.07.001_b15) 2018 Sun (10.1016/j.cag.2023.07.001_b34) 2020; 39 10.1016/j.cag.2023.07.001_b17 Butepage (10.1016/j.cag.2023.07.001_b53) 2017 Hernandez (10.1016/j.cag.2023.07.001_b55) 2019 Gui (10.1016/j.cag.2023.07.001_b56) 2018 Sofianos (10.1016/j.cag.2023.07.001_b7) 2021 Gao (10.1016/j.cag.2023.07.001_b39) 2016; 35 Zuffi (10.1016/j.cag.2023.07.001_b67) 2017 Aumentado-Armstrong (10.1016/j.cag.2023.07.001_b19) 2019 Ovsjanikov (10.1016/j.cag.2023.07.001_b73) 2012; 31 Mao (10.1016/j.cag.2023.07.001_b5) 2021; 129 Fragkiadaki (10.1016/j.cag.2023.07.001_b48) 2015 Guo (10.1016/j.cag.2023.07.001_b6) 2023 Chiu (10.1016/j.cag.2023.07.001_b51) 2019 Lehrmann (10.1016/j.cag.2023.07.001_b47) 2014 Fey (10.1016/j.cag.2023.07.001_b22) 2018 10.1016/j.cag.2023.07.001_b27 Taylor (10.1016/j.cag.2023.07.001_b46) 2006 Ranjan (10.1016/j.cag.2023.07.001_b32) 2018 Levie (10.1016/j.cag.2023.07.001_b33) 2019; 67 Tang (10.1016/j.cag.2023.07.001_b60) 2018 10.1016/j.cag.2023.07.001_b69 Marsot (10.1016/j.cag.2023.07.001_b9) 2021 Abrevaya (10.1016/j.cag.2023.07.001_b68) 2016; 58 Hartman (10.1016/j.cag.2023.07.001_b44) 2022 Community (10.1016/j.cag.2023.07.001_b72) 2018 Huang (10.1016/j.cag.2023.07.001_b26) 2021 Wang (10.1016/j.cag.2023.07.001_b45) 2005 Hanocka (10.1016/j.cag.2023.07.001_b25) 2019; 38 Welford (10.1016/j.cag.2023.07.001_b71) 1962; 4 Jain (10.1016/j.cag.2023.07.001_b49) 2016 Bruna (10.1016/j.cag.2023.07.001_b29) 2014 Mao (10.1016/j.cag.2023.07.001_b3) 2019 Rakotosaona (10.1016/j.cag.2023.07.001_b12) 2020 Li (10.1016/j.cag.2023.07.001_b2) 2018 Masci (10.1016/j.cag.2023.07.001_b20) 2015 Tan (10.1016/j.cag.2023.07.001_b40) 2017 Qi (10.1016/j.cag.2023.07.001_b18) 2017 Cai (10.1016/j.cag.2023.07.001_b62) 2020 Mao (10.1016/j.cag.2023.07.001_b4) 2020 Tevet (10.1016/j.cag.2023.07.001_b65) 2022 Loper (10.1016/j.cag.2023.07.001_b10) 2015; 34 Marin (10.1016/j.cag.2023.07.001_b36) 2021; 129 10.1016/j.cag.2023.07.001_b74 Guo (10.1016/j.cag.2023.07.001_b16) 2021; 43 Gong (10.1016/j.cag.2023.07.001_b13) 2019 Eisenberger (10.1016/j.cag.2023.07.001_b43) 2021 Aksan (10.1016/j.cag.2023.07.001_b61) 2021 Defferrard (10.1016/j.cag.2023.07.001_b30) 2016 Pegoraro (10.1016/j.cag.2023.07.001_b37) 2022; 41 Sohn (10.1016/j.cag.2023.07.001_b66) 2015 Lebailly (10.1016/j.cag.2023.07.001_b57) 2021 Gao (10.1016/j.cag.2023.07.001_b41) 2017 Marin (10.1016/j.cag.2023.07.001_b35) 2020 Ma (10.1016/j.cag.2023.07.001_b58) 2022 Lemeunier (10.1016/j.cag.2023.07.001_b14) 2022; 107 Boscaini (10.1016/j.cag.2023.07.001_b21) 2016 Vaswani (10.1016/j.cag.2023.07.001_b59) 2017 Bouritsas (10.1016/j.cag.2023.07.001_b24) 2019 Zhong (10.1016/j.cag.2023.07.001_b8) 2022 Kipf (10.1016/j.cag.2023.07.001_b31) 2017 Butepage (10.1016/j.cag.2023.07.001_b54) 2018 Bronstein (10.1016/j.cag.2023.07.001_b1) 2017; 34 |
| References_xml | – year: 2021 ident: b26 article-title: ARAPReg: An as-rigid-as possible regularization loss for learning deformable shape generators – year: 2021 ident: b61 article-title: A spatio-temporal transformer for 3D human motion prediction publication-title: 2021 International conference on 3D vision – start-page: 725 year: 2018 end-page: 741 ident: b32 article-title: Generating 3D faces using convolutional mesh autoencoders publication-title: Computer vision – ECCV 2018 – volume: 38 start-page: 1 year: 2019 end-page: 12 ident: b25 article-title: MeshCNN publication-title: ACM Trans Graph – year: 2017 ident: b31 article-title: Semi-supervised classification with graph convolutional networks publication-title: International conference on learning representations – volume: 41 start-page: 1 year: 2022 end-page: 16 ident: b38 article-title: DiffusionNet: Discretization agnostic learning on surfaces publication-title: ACM Trans Graph – year: 2015 ident: b48 article-title: Recurrent network models for human dynamics publication-title: 2015 IEEE international conference on computer vision – year: 2021 ident: b7 article-title: Space-time-separable graph convolutional network for pose forecasting publication-title: 2021 IEEE/CVF international conference on computer vision – year: 2023 ident: b6 article-title: Back to MLP: A simple baseline for human motion prediction publication-title: 2023 IEEE/CVF winter conference on applications of computer vision – year: 2022 ident: b58 article-title: Progressively generating better initial guesses towards next stages for high-quality human motion prediction publication-title: 2022 IEEE/CVF conference on computer vision and pattern recognition – year: 2021 ident: b63 article-title: Single-shot motion completion with transformer – year: 2020 ident: b35 article-title: Instant recovery of shape from spectrum via latent space connections publication-title: 2020 International conference on 3D vision – volume: 4 start-page: 419 year: 1962 end-page: 420 ident: b71 article-title: Note on a method for calculating corrected sums of squares and products publication-title: Technometrics – volume: 41 start-page: 13 year: 2022 end-page: 24 ident: b37 article-title: Localized shape modelling with global coherence: An inverse spectral approach publication-title: Comput Graph Forum – volume: 36 start-page: 1325 year: 2014 end-page: 1339 ident: b70 article-title: Human3.6M: Large scale datasets and predictive methods for 3D human sensing in natural environments publication-title: IEEE Trans Pattern Anal Mach Intell – year: 2019 ident: b55 article-title: Human motion prediction via spatio-temporal inpainting publication-title: 2019 IEEE/CVF international conference on computer vision – start-page: 5998 year: 2017 end-page: 6008 ident: b59 article-title: Attention is all you need publication-title: Advances in neural information processing systems – year: 2019 ident: b51 article-title: Action-agnostic human pose forecasting publication-title: 2019 IEEE winter conference on applications of computer vision – year: 2018 ident: b72 article-title: Blender - A 3D Modelling and Rendering Package – year: 2017 ident: b50 article-title: On human motion prediction using recurrent neural networks publication-title: 2017 IEEE conference on computer vision and pattern recognition – volume: 58 start-page: 12 year: 2016 end-page: 22 ident: b68 article-title: A 3D+t Laplace operator for temporal mesh sequences publication-title: Comput Graph – reference: Mahmood N, Ghorbani N, Troje NF, Pons-Moll G, Black MJ. AMASS: Archive of Motion Capture as Surface Shapes. In: International conference on computer vision. 2019, p. 5442–51. – volume: 67 start-page: 97 year: 2019 end-page: 109 ident: b33 article-title: CayleyNets: Graph convolutional neural networks with complex rational spectral filters publication-title: IEEE Trans Signal Process – year: 2014 ident: b47 article-title: Efficient nonlinear Markov models for human motion publication-title: 2014 IEEE conference on computer vision and pattern recognition – year: 2019 ident: b42 article-title: Mesh variational autoencoders with edge contraction pooling – year: 2020 ident: b28 article-title: Primal-dual mesh convolutional neural networks publication-title: Proceedings of the 34th international conference on neural information processing systems – year: 2021 ident: b43 article-title: NeuroMorph: Unsupervised shape interpolation and correspondence in one go publication-title: 2021 IEEE/CVF conference on computer vision and pattern recognition – start-page: 823 year: 2018 end-page: 842 ident: b56 article-title: Adversarial geometry-aware human motion prediction publication-title: Computer vision – ECCV 2018 – year: 2019 ident: b3 article-title: Learning trajectory dependencies for human motion prediction publication-title: 2019 IEEE/CVF international conference on computer vision – year: 2018 ident: b15 article-title: BERT: Pre-training of deep bidirectional transformers for language understanding – year: 2022 ident: b65 article-title: Human motion diffusion model – start-page: 474 year: 2020 end-page: 489 ident: b4 article-title: History repeats itself: Human motion prediction via motion attention publication-title: Computer vision – ECCV 2020 – year: 2016 ident: b49 article-title: Structural-RNN: Deep learning on spatio-temporal graphs publication-title: 2016 IEEE conference on computer vision and pattern recognition – year: 2018 ident: b2 article-title: Convolutional sequence to sequence model for human dynamics publication-title: 2018 IEEE/CVF conference on computer vision and pattern recognition – volume: 129 start-page: 2513 year: 2021 end-page: 2535 ident: b5 article-title: Multi-level motion attention for human motion prediction publication-title: Int J Comput Vis – start-page: 651 year: 2021 end-page: 665 ident: b57 article-title: Motion prediction using temporal inception module publication-title: Computer vision – ACCV 2020 – year: 2017 ident: b53 article-title: Deep representation learning for human motion prediction and classification publication-title: 2017 IEEE conference on computer vision and pattern recognition – year: 2021 ident: b9 article-title: A structured latent space for human body motion generation – year: 2018 ident: b22 article-title: SplineCNN: Fast geometric deep learning with continuous B-spline kernels publication-title: 2018 IEEE/CVF conference on computer vision and pattern recognition – start-page: 226 year: 2020 end-page: 242 ident: b62 article-title: Learning progressive joint propagation for human motion prediction publication-title: Computer vision – ECCV 2020 – year: 2017 ident: b67 article-title: 3D menagerie: Modeling the 3D shape and pose of animals publication-title: IEEE conf. on computer vision and pattern recognition – volume: 31 start-page: 1 year: 2012 end-page: 11 ident: b73 article-title: Functional maps publication-title: ACM Trans Graph – year: 2022 ident: b8 article-title: Spatio-temporal gating-adjacency GCN for human motion prediction publication-title: 2022 IEEE/CVF conference on computer vision and pattern recognition – year: 2006 ident: b46 article-title: Modeling human motion using binary latent variables publication-title: NIPS – volume: 43 start-page: 4338 year: 2021 end-page: 4364 ident: b16 article-title: Deep learning for 3D point clouds: A survey publication-title: IEEE Trans Pattern Anal Mach Intell – reference: Qi CR, Su H, Mo K, Guibas LJ. PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation. In: Proceedings of the IEEE conference on computer vision and pattern recognition. CVPR, 2017. – start-page: 19 year: 2020 end-page: 35 ident: b11 article-title: LIMP: Learning latent shape representations with metric preservation priors publication-title: Computer vision – ECCV 2020 – year: 2017 ident: b41 article-title: Sparse data driven mesh deformation – year: 2014 ident: b29 article-title: Spectral networks and locally connected networks on graphs publication-title: International conference on learning representations (ICLR2014), CBLS, April 2014 – volume: 129 start-page: 2745 year: 2021 end-page: 2760 ident: b36 article-title: Spectral shape recovery and analysis via data-driven connections publication-title: Int J Comput Vis – start-page: 655 year: 2020 end-page: 672 ident: b12 article-title: Intrinsic point cloud interpolation via dual latent space navigation publication-title: Computer vision – ECCV 2020 – start-page: 3197 year: 2016 end-page: 3205 ident: b21 article-title: Learning shape correspondence with anisotropic convolutional neural networks publication-title: Proceedings of the 30th international conference on neural information processing systems – volume: 39 start-page: 204 year: 2020 end-page: 216 ident: b34 article-title: ZerNet: Convolutional neural networks on arbitrary surfaces via zernike local tangent space estimation publication-title: Comput Graph Forum – year: 2022 ident: b44 article-title: Bare-ESA: A Riemannian framework for unregistered human body shapes – volume: 34 start-page: 1 year: 2015 end-page: 16 ident: b10 article-title: SMPL publication-title: ACM Trans Graph – year: 2019 ident: b13 article-title: SpiralNet++: A fast and highly efficient mesh convolution operator publication-title: 2019 IEEE/CVF international conference on computer vision workshop – year: 2019 ident: b52 article-title: Towards natural and accurate future motion prediction of humans and animals publication-title: 2019 IEEE/CVF conference on computer vision and pattern recognition – year: 2015 ident: b20 article-title: Geodesic convolutional neural networks on Riemannian manifolds publication-title: 2015 IEEE international conference on computer vision workshop – year: 2005 ident: b45 article-title: Gaussian process dynamical models publication-title: Advances in neural information processing systems, Vol. 18 – start-page: 935 year: 2018 end-page: 941 ident: b60 article-title: Long-term human motion prediction by modeling motion context and enhancing motion dynamic publication-title: Proceedings of the 27th international joint conference on artificial intelligence – volume: 35 start-page: 1 year: 2016 end-page: 17 ident: b39 article-title: Efficient and flexible deformation representation for data-driven surface modeling publication-title: ACM Trans Graph – start-page: 480 year: 2022 end-page: 497 ident: b64 article-title: TEMOS: Generating diverse human motions from textual descriptions publication-title: Lecture notes in computer science – volume: 34 start-page: 18 year: 2017 end-page: 42 ident: b1 article-title: Geometric deep learning: Going beyond euclidean data publication-title: IEEE Signal Process Mag – year: 2019 ident: b24 article-title: Neural 3D morphable models: Spiral convolutional networks for 3D shape representation learning and generation publication-title: 2019 IEEE/CVF international conference on computer vision – start-page: 3844 year: 2016 end-page: 3852 ident: b30 article-title: Convolutional neural networks on graphs with fast localized spectral filtering publication-title: Proceedings of the 30th international conference on neural information processing systems – year: 2017 ident: b40 article-title: Variational autoencoders for deforming 3D mesh models – start-page: 5105 year: 2017 end-page: 5114 ident: b18 article-title: PointNet++: Deep hierarchical feature learning on point sets in a metric space publication-title: Proceedings of the 31st international conference on neural information processing systems – year: 2018 ident: b54 article-title: Anticipating many futures: Online human motion prediction and generation for human-robot interaction publication-title: 2018 IEEE international conference on robotics and automation – start-page: 3483 year: 2015 end-page: 3491 ident: b66 article-title: Learning structured output representation using deep conditional generative models publication-title: Advances in neural information processing systems – reference: Verma N, Boyer E, Verbeek J. FeaStNet: Feature-Steered Graph Convolutions for 3D Shape Analysis. In: CVPR - IEEE conference on computer vision & pattern recognition. Salt Lake City, United States; 2018, URL. – reference: Pavlakos G, Choutas V, Ghorbani N, Bolkart T, Osman AAA, Tzionas D, et al. Expressive Body Capture: 3D Hands, Face, and Body from a Single Image. In: Proceedings IEEE conf. on computer vision and pattern recognition. CVPR, 2019, p. 10975–85. – year: 2019 ident: b19 article-title: Geometric disentanglement for generative latent shape models publication-title: 2019 IEEE/CVF international conference on computer vision – start-page: 349 year: 2019 end-page: 362 ident: b23 article-title: A simple approach to intrinsic correspondence learning on unstructured 3D meshes publication-title: Lecture notes in computer science – volume: 107 start-page: 131 year: 2022 end-page: 143 ident: b14 article-title: Representation learning of 3D meshes using an autoencoder in the spectral domain publication-title: Comput Graph – year: 2018 ident: 10.1016/j.cag.2023.07.001_b22 article-title: SplineCNN: Fast geometric deep learning with continuous B-spline kernels – start-page: 3844 year: 2016 ident: 10.1016/j.cag.2023.07.001_b30 article-title: Convolutional neural networks on graphs with fast localized spectral filtering – year: 2005 ident: 10.1016/j.cag.2023.07.001_b45 article-title: Gaussian process dynamical models – year: 2022 ident: 10.1016/j.cag.2023.07.001_b44 – year: 2017 ident: 10.1016/j.cag.2023.07.001_b41 – year: 2017 ident: 10.1016/j.cag.2023.07.001_b53 article-title: Deep representation learning for human motion prediction and classification – year: 2019 ident: 10.1016/j.cag.2023.07.001_b3 article-title: Learning trajectory dependencies for human motion prediction – ident: 10.1016/j.cag.2023.07.001_b69 doi: 10.1109/ICCV.2019.00554 – volume: 4 start-page: 419 issue: 3 year: 1962 ident: 10.1016/j.cag.2023.07.001_b71 article-title: Note on a method for calculating corrected sums of squares and products publication-title: Technometrics doi: 10.1080/00401706.1962.10490022 – volume: 39 start-page: 204 issue: 6 year: 2020 ident: 10.1016/j.cag.2023.07.001_b34 article-title: ZerNet: Convolutional neural networks on arbitrary surfaces via zernike local tangent space estimation publication-title: Comput Graph Forum doi: 10.1111/cgf.14012 – year: 2014 ident: 10.1016/j.cag.2023.07.001_b29 article-title: Spectral networks and locally connected networks on graphs – year: 2019 ident: 10.1016/j.cag.2023.07.001_b24 article-title: Neural 3D morphable models: Spiral convolutional networks for 3D shape representation learning and generation – year: 2021 ident: 10.1016/j.cag.2023.07.001_b9 – year: 2019 ident: 10.1016/j.cag.2023.07.001_b42 – ident: 10.1016/j.cag.2023.07.001_b27 doi: 10.1109/CVPR.2018.00275 – volume: 67 start-page: 97 issue: 1 year: 2019 ident: 10.1016/j.cag.2023.07.001_b33 article-title: CayleyNets: Graph convolutional neural networks with complex rational spectral filters publication-title: IEEE Trans Signal Process doi: 10.1109/TSP.2018.2879624 – start-page: 5105 year: 2017 ident: 10.1016/j.cag.2023.07.001_b18 article-title: PointNet++: Deep hierarchical feature learning on point sets in a metric space – volume: 129 start-page: 2745 issue: 10 year: 2021 ident: 10.1016/j.cag.2023.07.001_b36 article-title: Spectral shape recovery and analysis via data-driven connections publication-title: Int J Comput Vis doi: 10.1007/s11263-021-01492-6 – start-page: 3197 year: 2016 ident: 10.1016/j.cag.2023.07.001_b21 article-title: Learning shape correspondence with anisotropic convolutional neural networks – year: 2021 ident: 10.1016/j.cag.2023.07.001_b26 – year: 2015 ident: 10.1016/j.cag.2023.07.001_b48 article-title: Recurrent network models for human dynamics – year: 2021 ident: 10.1016/j.cag.2023.07.001_b63 – year: 2022 ident: 10.1016/j.cag.2023.07.001_b65 – start-page: 5998 year: 2017 ident: 10.1016/j.cag.2023.07.001_b59 article-title: Attention is all you need – volume: 34 start-page: 18 issue: 4 year: 2017 ident: 10.1016/j.cag.2023.07.001_b1 article-title: Geometric deep learning: Going beyond euclidean data publication-title: IEEE Signal Process Mag doi: 10.1109/MSP.2017.2693418 – ident: 10.1016/j.cag.2023.07.001_b74 doi: 10.1109/CVPR.2019.01123 – year: 2019 ident: 10.1016/j.cag.2023.07.001_b19 article-title: Geometric disentanglement for generative latent shape models – year: 2017 ident: 10.1016/j.cag.2023.07.001_b50 article-title: On human motion prediction using recurrent neural networks – year: 2018 ident: 10.1016/j.cag.2023.07.001_b15 – year: 2006 ident: 10.1016/j.cag.2023.07.001_b46 article-title: Modeling human motion using binary latent variables – start-page: 725 year: 2018 ident: 10.1016/j.cag.2023.07.001_b32 article-title: Generating 3D faces using convolutional mesh autoencoders doi: 10.1007/978-3-030-01219-9_43 – year: 2019 ident: 10.1016/j.cag.2023.07.001_b55 article-title: Human motion prediction via spatio-temporal inpainting – volume: 36 start-page: 1325 issue: 7 year: 2014 ident: 10.1016/j.cag.2023.07.001_b70 article-title: Human3.6M: Large scale datasets and predictive methods for 3D human sensing in natural environments publication-title: IEEE Trans Pattern Anal Mach Intell doi: 10.1109/TPAMI.2013.248 – start-page: 349 year: 2019 ident: 10.1016/j.cag.2023.07.001_b23 article-title: A simple approach to intrinsic correspondence learning on unstructured 3D meshes doi: 10.1007/978-3-030-11015-4_26 – year: 2017 ident: 10.1016/j.cag.2023.07.001_b40 – volume: 41 start-page: 13 issue: 5 year: 2022 ident: 10.1016/j.cag.2023.07.001_b37 article-title: Localized shape modelling with global coherence: An inverse spectral approach publication-title: Comput Graph Forum doi: 10.1111/cgf.14599 – volume: 43 start-page: 4338 issue: 12 year: 2021 ident: 10.1016/j.cag.2023.07.001_b16 article-title: Deep learning for 3D point clouds: A survey publication-title: IEEE Trans Pattern Anal Mach Intell doi: 10.1109/TPAMI.2020.3005434 – year: 2021 ident: 10.1016/j.cag.2023.07.001_b61 article-title: A spatio-temporal transformer for 3D human motion prediction – year: 2015 ident: 10.1016/j.cag.2023.07.001_b20 article-title: Geodesic convolutional neural networks on Riemannian manifolds – volume: 35 start-page: 1 issue: 5 year: 2016 ident: 10.1016/j.cag.2023.07.001_b39 article-title: Efficient and flexible deformation representation for data-driven surface modeling publication-title: ACM Trans Graph doi: 10.1145/2908736 – year: 2017 ident: 10.1016/j.cag.2023.07.001_b67 article-title: 3D menagerie: Modeling the 3D shape and pose of animals – year: 2021 ident: 10.1016/j.cag.2023.07.001_b43 article-title: NeuroMorph: Unsupervised shape interpolation and correspondence in one go – start-page: 655 year: 2020 ident: 10.1016/j.cag.2023.07.001_b12 article-title: Intrinsic point cloud interpolation via dual latent space navigation – volume: 38 start-page: 1 issue: 4 year: 2019 ident: 10.1016/j.cag.2023.07.001_b25 article-title: MeshCNN publication-title: ACM Trans Graph doi: 10.1145/3306346.3322959 – start-page: 226 year: 2020 ident: 10.1016/j.cag.2023.07.001_b62 article-title: Learning progressive joint propagation for human motion prediction – year: 2014 ident: 10.1016/j.cag.2023.07.001_b47 article-title: Efficient nonlinear Markov models for human motion – year: 2018 ident: 10.1016/j.cag.2023.07.001_b2 article-title: Convolutional sequence to sequence model for human dynamics – year: 2023 ident: 10.1016/j.cag.2023.07.001_b6 article-title: Back to MLP: A simple baseline for human motion prediction – volume: 129 start-page: 2513 issue: 9 year: 2021 ident: 10.1016/j.cag.2023.07.001_b5 article-title: Multi-level motion attention for human motion prediction publication-title: Int J Comput Vis doi: 10.1007/s11263-021-01483-7 – volume: 34 start-page: 1 issue: 6 year: 2015 ident: 10.1016/j.cag.2023.07.001_b10 article-title: SMPL publication-title: ACM Trans Graph doi: 10.1145/2816795.2818013 – start-page: 823 year: 2018 ident: 10.1016/j.cag.2023.07.001_b56 article-title: Adversarial geometry-aware human motion prediction – volume: 58 start-page: 12 year: 2016 ident: 10.1016/j.cag.2023.07.001_b68 article-title: A 3D+t Laplace operator for temporal mesh sequences publication-title: Comput Graph doi: 10.1016/j.cag.2016.05.018 – year: 2020 ident: 10.1016/j.cag.2023.07.001_b28 article-title: Primal-dual mesh convolutional neural networks – start-page: 19 year: 2020 ident: 10.1016/j.cag.2023.07.001_b11 article-title: LIMP: Learning latent shape representations with metric preservation priors – volume: 107 start-page: 131 year: 2022 ident: 10.1016/j.cag.2023.07.001_b14 article-title: Representation learning of 3D meshes using an autoencoder in the spectral domain publication-title: Comput Graph doi: 10.1016/j.cag.2022.07.011 – volume: 41 start-page: 1 issue: 3 year: 2022 ident: 10.1016/j.cag.2023.07.001_b38 article-title: DiffusionNet: Discretization agnostic learning on surfaces publication-title: ACM Trans Graph doi: 10.1145/3507905 – year: 2019 ident: 10.1016/j.cag.2023.07.001_b51 article-title: Action-agnostic human pose forecasting – start-page: 474 year: 2020 ident: 10.1016/j.cag.2023.07.001_b4 article-title: History repeats itself: Human motion prediction via motion attention – start-page: 3483 year: 2015 ident: 10.1016/j.cag.2023.07.001_b66 article-title: Learning structured output representation using deep conditional generative models – start-page: 651 year: 2021 ident: 10.1016/j.cag.2023.07.001_b57 article-title: Motion prediction using temporal inception module – year: 2021 ident: 10.1016/j.cag.2023.07.001_b7 article-title: Space-time-separable graph convolutional network for pose forecasting – ident: 10.1016/j.cag.2023.07.001_b17 – year: 2016 ident: 10.1016/j.cag.2023.07.001_b49 article-title: Structural-RNN: Deep learning on spatio-temporal graphs – year: 2017 ident: 10.1016/j.cag.2023.07.001_b31 article-title: Semi-supervised classification with graph convolutional networks – year: 2019 ident: 10.1016/j.cag.2023.07.001_b52 article-title: Towards natural and accurate future motion prediction of humans and animals – year: 2022 ident: 10.1016/j.cag.2023.07.001_b58 article-title: Progressively generating better initial guesses towards next stages for high-quality human motion prediction – volume: 31 start-page: 1 issue: 4 year: 2012 ident: 10.1016/j.cag.2023.07.001_b73 article-title: Functional maps publication-title: ACM Trans Graph doi: 10.1145/2185520.2185526 – year: 2020 ident: 10.1016/j.cag.2023.07.001_b35 article-title: Instant recovery of shape from spectrum via latent space connections – start-page: 935 year: 2018 ident: 10.1016/j.cag.2023.07.001_b60 article-title: Long-term human motion prediction by modeling motion context and enhancing motion dynamic doi: 10.24963/ijcai.2018/130 – start-page: 480 year: 2022 ident: 10.1016/j.cag.2023.07.001_b64 article-title: TEMOS: Generating diverse human motions from textual descriptions doi: 10.1007/978-3-031-20047-2_28 – year: 2022 ident: 10.1016/j.cag.2023.07.001_b8 article-title: Spatio-temporal gating-adjacency GCN for human motion prediction – year: 2019 ident: 10.1016/j.cag.2023.07.001_b13 article-title: SpiralNet++: A fast and highly efficient mesh convolution operator – year: 2018 ident: 10.1016/j.cag.2023.07.001_b72 – year: 2018 ident: 10.1016/j.cag.2023.07.001_b54 article-title: Anticipating many futures: Online human motion prediction and generation for human-robot interaction |
| SSID | ssj0002264 |
| Score | 2.393756 |
| Snippet | We present SpecTrHuMS, a Spectral Transformer for 3D triangular Human Mesh Sequence learning which combines known deep learning models with spectral mesh... |
| SourceID | hal crossref elsevier |
| SourceType | Open Access Repository Enrichment Source Index Database Publisher |
| StartPage | 191 |
| SubjectTerms | Autoencoder Computer Science Computer Vision and Pattern Recognition Geometric deep learning Human body triangular mesh sequences Machine Learning Spectral analysis Transformers |
| Title | SpecTrHuMS: Spectral transformer for human mesh sequence learning |
| URI | https://dx.doi.org/10.1016/j.cag.2023.07.001 https://hal.science/hal-04154623 |
| Volume | 115 |
| WOSCitedRecordID | wos001056807800001&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: 0097-8493 databaseCode: AIEXJ dateStart: 19950101 customDbUrl: isFulltext: true dateEnd: 99991231 titleUrlDefault: https://www.sciencedirect.com omitProxy: false ssIdentifier: ssj0002264 providerName: Elsevier |
| link | http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwtV3db9MwELdKxwM8THxqY4AixBMoKG7i2OGtYpsKKhMSBfUtshNnH2rTqm2qvfC_c2fHXhjaxB54SRs3dtLcT747--53hLzlCStoVJYhk0kaJrRMQwlWCBhymUT-KCVMEM3PMT85EdNp9q3X--VyYbYzXtfi8jJb_ldRQxsIG1Nn7yBuPyg0wHcQOhxB7HD8J8FjRfnJatR8_Y7ePp7hYgbWgrAWql6Z0EJbnG-u12fvXTi1KyFx2rVYXdmHtQGJ4bfuBMiP9Vw39XmbTTiz2-7dYJpDXVsWg-PZwiQW-p5yu2js9WZtvsHyR83c_37YLBeWE8H23HTXJwZXkW5-zs1ADya2DqKfcynrzJrUFuxqFfDAkB78PbfbZYYL8NtPP-CNDOlqe6c_eLSv6TcfdegC2i5yGCLHIfIId9_Bed4ZcJaJPtkZfj6afvGqHLOMLY2p_QtuW9wECF57jpsMm3tnbonemCyTR2S39TWCocXIY9LT9RPysMNA-RS0qkfLx8BhJehgJYCPwGAlQKwEDiuBw8oz8uP4aPJpFLZVNcIi5skmVIViQupCcaErJVlcRmUcpxlm0ZdlxBSrmK4KUMiUVkjlI3icKCrjgaQlRlE9J_16Ues9EihNJU0zEXHNkrQSWYHlgySveAmGrxL7JHKvJC9aynmsfDLLbxTFPnnnuywt38ptFyfuPeetwWgNwRwwc1u3NyATPzwSrI-G4xzbkLAiAY9gS1_c5UEOyIMr6L8k_c2q0a_I_WK7OV-vXreo-g1Ol5OD |
| 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=SpecTrHuMS%3A+Spectral+transformer+for+human+mesh+sequence+learning&rft.jtitle=Computers+%26+graphics&rft.au=Lemeunier%2C+Cl%C3%A9ment&rft.au=Denis%2C+Florence&rft.au=Lavou%C3%A9%2C+Guillaume&rft.au=Dupont%2C+Florent&rft.date=2023-10-01&rft.issn=0097-8493&rft.volume=115&rft.spage=191&rft.epage=203&rft_id=info:doi/10.1016%2Fj.cag.2023.07.001&rft.externalDBID=n%2Fa&rft.externalDocID=10_1016_j_cag_2023_07_001 |
| thumbnail_l | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=0097-8493&client=summon |
| thumbnail_m | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=0097-8493&client=summon |
| thumbnail_s | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=0097-8493&client=summon |