D-IF: Uncertainty-aware Human Digitization via Implicit Distribution Field
Realistic virtual humans play a crucial role in numerous industries, such as metaverse, intelligent healthcare, and self-driving simulation. But creating them on a large scale with high levels of realism remains a challenge. The utilization of deep implicit function sparks a new era of image-based 3...
Saved in:
Published in | 2023 IEEE/CVF International Conference on Computer Vision (ICCV) pp. 9088 - 9098 |
---|---|
Main Authors | , , , , , |
Format | Conference Proceeding |
Language | English |
Published |
IEEE
01.10.2023
|
Subjects | |
Online Access | Get full text |
Cover
Loading…
Abstract | Realistic virtual humans play a crucial role in numerous industries, such as metaverse, intelligent healthcare, and self-driving simulation. But creating them on a large scale with high levels of realism remains a challenge. The utilization of deep implicit function sparks a new era of image-based 3D clothed human reconstruction, enabling pixel-aligned shape recovery with fine details. Subsequently, the vast majority of works locate the surface by regressing the deterministic implicit value for each point. However, should all points be treated equally regardless of their proximity to the surface? In this paper, we propose replacing the implicit value with an adaptive uncertainty distribution, to differentiate between points based on their distance to the surface. This simple "value ⇒ distribution" transition yields significant improvements on nearly all the baselines. Furthermore, qualitative results demonstrate that the models trained using our uncertainty distribution loss, can capture more intricate wrinkles, and realistic limbs. Code and models are available for research purposes at github.com/psyai-net/D-IF release. |
---|---|
AbstractList | Realistic virtual humans play a crucial role in numerous industries, such as metaverse, intelligent healthcare, and self-driving simulation. But creating them on a large scale with high levels of realism remains a challenge. The utilization of deep implicit function sparks a new era of image-based 3D clothed human reconstruction, enabling pixel-aligned shape recovery with fine details. Subsequently, the vast majority of works locate the surface by regressing the deterministic implicit value for each point. However, should all points be treated equally regardless of their proximity to the surface? In this paper, we propose replacing the implicit value with an adaptive uncertainty distribution, to differentiate between points based on their distance to the surface. This simple "value ⇒ distribution" transition yields significant improvements on nearly all the baselines. Furthermore, qualitative results demonstrate that the models trained using our uncertainty distribution loss, can capture more intricate wrinkles, and realistic limbs. Code and models are available for research purposes at github.com/psyai-net/D-IF release. |
Author | Xu, Hao Wang, Wei Fan, Zhaoxin Xiu, Yuliang Yang, Xueting Luo, Yihao |
Author_xml | – sequence: 1 givenname: Xueting surname: Yang fullname: Yang, Xueting email: yangxueting@psyai.net organization: Psyche AI Inc – sequence: 2 givenname: Yihao surname: Luo fullname: Luo, Yihao email: y.luo23@imperial.ac.uk organization: Psyche AI Inc – sequence: 3 givenname: Yuliang surname: Xiu fullname: Xiu, Yuliang email: yuliang.xiu@tue.mpg.de organization: Max Planck Institute for Intelligent Systems – sequence: 4 givenname: Wei surname: Wang fullname: Wang, Wei email: faithwwei@bupt.edu.cn organization: Psyche AI Inc – sequence: 5 givenname: Hao surname: Xu fullname: Xu, Hao email: hxubl@connect.ust.hk organization: Psyche AI Inc – sequence: 6 givenname: Zhaoxin surname: Fan fullname: Fan, Zhaoxin email: zfanaq@connect.ust.hk organization: Psyche AI Inc |
BookMark | eNotzMFOwkAQgOHVaCIgb8ChL1Ccnd12dr2ZIlJD4kW8kt12a8aUQtpFg0-PUU__4Uv-sbjq9l0QYiZhLiXYu7Io3jIJBHMEVHMAo-hCTC1ZozJQQNLoSzFCZSClDPSNGA_DB4CyaPKReF6k5fI-2XRV6KPjLp5S9-X6kKyOO9clC37nyN8u8r5LPtkl5e7QcsXxR4bYsz_-ypJDW9-K68a1Q5j-dyI2y8fXYpWuX57K4mGdMoKOKRLqCmQO1nnr8sZg7mtNXlck61xhbQizukFyCkn73DfWaYvBUoOorVcTMfv7cghhe-h55_rTVoIisplWZ6AfTq4 |
CODEN | IEEPAD |
ContentType | Conference Proceeding |
DBID | 6IE 6IH CBEJK RIE RIO |
DOI | 10.1109/ICCV51070.2023.00837 |
DatabaseName | IEEE Electronic Library (IEL) Conference Proceedings IEEE Proceedings Order Plan (POP) 1998-present by volume IEEE Xplore All Conference Proceedings IEL IEEE Proceedings Order Plans (POP) 1998-present |
DatabaseTitleList | |
Database_xml | – sequence: 1 dbid: RIE name: IEEE Electronic Library (IEL) url: https://proxy.k.utb.cz/login?url=https://ieeexplore.ieee.org/ sourceTypes: Publisher |
DeliveryMethod | fulltext_linktorsrc |
Discipline | Applied Sciences |
EISBN | 9798350307184 |
EISSN | 2380-7504 |
EndPage | 9098 |
ExternalDocumentID | 10377954 |
Genre | orig-research |
GroupedDBID | 29O 6IE 6IF 6IH 6IK 6IL 6IM 6IN AAJGR ACGFS ADZIZ ALMA_UNASSIGNED_HOLDINGS BEFXN BFFAM BGNUA BKEBE BPEOZ CBEJK CHZPO IPLJI JC5 M43 OCL RIE RIL RIO RNS |
ID | FETCH-LOGICAL-i204t-2724c01609ab9a6f826bd47b4c71d632d8725df27a3274b6bf9a492e97f2249b3 |
IEDL.DBID | RIE |
IngestDate | Wed Jun 26 19:28:16 EDT 2024 |
IsPeerReviewed | false |
IsScholarly | true |
Language | English |
LinkModel | DirectLink |
MergedId | FETCHMERGED-LOGICAL-i204t-2724c01609ab9a6f826bd47b4c71d632d8725df27a3274b6bf9a492e97f2249b3 |
PageCount | 11 |
ParticipantIDs | ieee_primary_10377954 |
PublicationCentury | 2000 |
PublicationDate | 2023-Oct.-1 |
PublicationDateYYYYMMDD | 2023-10-01 |
PublicationDate_xml | – month: 10 year: 2023 text: 2023-Oct.-1 day: 01 |
PublicationDecade | 2020 |
PublicationTitle | 2023 IEEE/CVF International Conference on Computer Vision (ICCV) |
PublicationTitleAbbrev | ICCV |
PublicationYear | 2023 |
Publisher | IEEE |
Publisher_xml | – name: IEEE |
SSID | ssj0039286 |
Score | 2.3310106 |
Snippet | Realistic virtual humans play a crucial role in numerous industries, such as metaverse, intelligent healthcare, and self-driving simulation. But creating them... |
SourceID | ieee |
SourceType | Publisher |
StartPage | 9088 |
SubjectTerms | Medical services Metaverse Shape Sparks Surface reconstruction Three-dimensional displays Uncertainty |
Title | D-IF: Uncertainty-aware Human Digitization via Implicit Distribution Field |
URI | https://ieeexplore.ieee.org/document/10377954 |
hasFullText | 1 |
inHoldings | 1 |
isFullTextHit | |
isPrint | |
link | http://utb.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwjV3PT8MgGCVuJ0_zx4y_w8ErXQcUitfNZlvi4sGZ3RYoYBqTzWytRv96gXYzmph4a8uBBvrx-Oh77wPghlFp-3kikGtniGKlXEhRgojFqaZKYhXc9e-nbDSjk3kyb8TqQQtjjAnkMxP5y_AvX6_yyh-V9bymjYuEtkArjXEt1touuw7nU9Zo4_qx6I0Hgyf3vfE48gXCvYsp-VlBJQBI1gHTbdc1b-QlqkoV5Z-_XBn__W4HoPut1YMPOxQ6BHtmeQQ6zeYSNqG7OQaTIRpnt3Dm7gIJoPxA8l2uDQzH-HBYPBdlo8mEb4WE40A1L0rXstlVxYKZJ7x1wSy7exyMUFNIARU4piXCHNPcW8kJqYRk1qUUSlOuaM77mhGsU44TbTGXxCWpiikrJBXYCG4dwgtFTkB7uVqaUwCTNMcyES5vNtqlnkK4-E4sYX7fZ21Kz0DXj83itfbKWGyH5fyP5xdg389PTY-7BO1yXZkrB_Olug7T-wVzP6Xw |
link.rule.ids | 310,311,786,790,795,796,802,27958,55109 |
linkProvider | IEEE |
linkToHtml | http://utb.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwjV3PT8IwGG0UD3rCHxh_24PXDmi7dvUKLgyBeADDjbRbaxYTMDA0-tfbdkOjiYm3bd-SLW2-vX7de-8D4IZRadppKJCNM0SxUjalKEHE4CijSmLl3fWHI9ab0P40nFZida-F0Vp78pkO3KH_l58t0rXbKms6TRsXId0GOxboW6KUa20-vBbpI1ap42ywmXQ6j_Yu3gpci3DnY0p-9lDxEBLXwWjz8JI58hysCxWkH798Gf_9dvug8a3Wgw9fOHQAtvT8ENSr5SWsknd1BPpdlMS3cGLPPA2geEfyTS419Bv5sJs_5UWlyoSvuYSJJ5vnhY2svvpiwdhR3hpgEt-NOz1UtVJAOW7RAmGOaerM5IRUQjJjiwqVUa5oytsZIziLOA4zg7kktkxVTBkhqcBacGMxXihyDGrzxVyfABhGKZahsJWzzmzxKYTN8NAQ5lZ-xkT0FDTc2MxeSreM2WZYzv64fg12e-PhYDZIRvfnYM_NVUmWuwC1YrnWlxb0C3Xlp_oTaGepRg |
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%3Abook&rft.genre=proceeding&rft.title=2023+IEEE%2FCVF+International+Conference+on+Computer+Vision+%28ICCV%29&rft.atitle=D-IF%3A+Uncertainty-aware+Human+Digitization+via+Implicit+Distribution+Field&rft.au=Yang%2C+Xueting&rft.au=Luo%2C+Yihao&rft.au=Xiu%2C+Yuliang&rft.au=Wang%2C+Wei&rft.date=2023-10-01&rft.pub=IEEE&rft.eissn=2380-7504&rft.spage=9088&rft.epage=9098&rft_id=info:doi/10.1109%2FICCV51070.2023.00837&rft.externalDocID=10377954 |