서브메뉴
검색
Reconstructing and Designing 3D Objects Across the Physical and Digital Worlds
Reconstructing and Designing 3D Objects Across the Physical and Digital Worlds
상세정보
- 자료유형
- 학위논문 서양
- 최종처리일시
- 20260202105628
- ISBN
- 9798265430113
- DDC
- 574
- 저자명
- Guo, Michelle.
- 서명/저자
- Reconstructing and Designing 3D Objects Across the Physical and Digital Worlds
- 발행사항
- [Sl] : Stanford University, 2025
- 발행사항
- Ann Arbor : ProQuest Dissertations & Theses, 2025
- 형태사항
- 112 p
- 주기사항
- Source: Dissertations Abstracts International, Volume: 87-05, Section: A.
- 주기사항
- Advisor: Liu, Cheng-Yun;Wu, Jiajun.
- 학위논문주기
- Thesis (Ph.D.)--Stanford University, 2025.
- 초록/해제
- 요약The integration of the digital and physical worlds requires the ability to accurately reconstruct and design three-dimensional objects. Current approaches often struggle to (i) reproduce the appearance and behavior of real-world objects and (ii) create object designs that are functional for humans and robots. This dissertation presents frameworks across the two lines of research.The first line of research focuses on reconstruction. I present a neural object representation capable of reconstructing object appearance from images, handling complex materials (both opaque and translucent), and supporting free-viewpoint relighting and scene composition. I also introduce a method for reconstructing photorealistic, simulation-ready garments from a single multi-view capture, using a hybrid mesh-embedded 3D Gaussian splat representation. Such a representation allows simulation-readiness and generalization to novel poses and lighting conditions.The second line of research explores 3D object design. This includes a mesh deformation framework for generating objects that follow both semantic guidance and contact constraints. Furthermore, I propose a learning-based framework for automatically designing 3D printable adaptations on everyday objects for robot manipulation, formulating adaptation design and control as a dual Markov Decision Process to improve "robot ergonomics" for challenging tasks.Together, these contributions advance the state-of-the-art in digital twins and functional object design, enabling more robust agent-object interactions in both the virtual and physical worlds.
- 일반주제명
- Adaptation
- 일반주제명
- Design
- 일반주제명
- Lighting
- 일반주제명
- Semantics
- 기타저자
- Stanford University.
- 기본자료저록
- Dissertations Abstracts International. 87-05A.
- 전자적 위치 및 접속
- 로그인 후 원문을 볼 수 있습니다.
MARC
008260126s2025 us c eng d■001000017360851
■00520260202105628
■006m o d
■007cr#unu||||||||
■020 ▼a9798265430113
■035 ▼a(MiAaPQ)AAI32316576
■035 ▼a(MiAaPQ)Stanfordpt199vf1217
■040 ▼aMiAaPQ▼cMiAaPQ
■0820 ▼a574
■1001 ▼aGuo, Michelle.
■24510▼aReconstructing and Designing 3D Objects Across the Physical and Digital Worlds
■260 ▼a[Sl]▼bStanford University▼c2025
■260 1▼aAnn Arbor▼bProQuest Dissertations & Theses▼c2025
■300 ▼a112 p
■500 ▼aSource: Dissertations Abstracts International, Volume: 87-05, Section: A.
■500 ▼aAdvisor: Liu, Cheng-Yun;Wu, Jiajun.
■5021 ▼aThesis (Ph.D.)--Stanford University, 2025.
■520 ▼aThe integration of the digital and physical worlds requires the ability to accurately reconstruct and design three-dimensional objects. Current approaches often struggle to (i) reproduce the appearance and behavior of real-world objects and (ii) create object designs that are functional for humans and robots. This dissertation presents frameworks across the two lines of research.The first line of research focuses on reconstruction. I present a neural object representation capable of reconstructing object appearance from images, handling complex materials (both opaque and translucent), and supporting free-viewpoint relighting and scene composition. I also introduce a method for reconstructing photorealistic, simulation-ready garments from a single multi-view capture, using a hybrid mesh-embedded 3D Gaussian splat representation. Such a representation allows simulation-readiness and generalization to novel poses and lighting conditions.The second line of research explores 3D object design. This includes a mesh deformation framework for generating objects that follow both semantic guidance and contact constraints. Furthermore, I propose a learning-based framework for automatically designing 3D printable adaptations on everyday objects for robot manipulation, formulating adaptation design and control as a dual Markov Decision Process to improve "robot ergonomics" for challenging tasks.Together, these contributions advance the state-of-the-art in digital twins and functional object design, enabling more robust agent-object interactions in both the virtual and physical worlds.
■590 ▼aSchool code: 0212.
■650 4▼aAdaptation
■650 4▼aDesign
■650 4▼aLighting
■650 4▼aSemantics
■690 ▼a0389
■71020▼aStanford University.
■7730 ▼tDissertations Abstracts International▼g87-05A.
■790 ▼a0212
■791 ▼aPh.D.
■792 ▼a2025
■793 ▼aEnglish
■85640▼uhttp://www.riss.kr/pdu/ddodLink.do?id=T17360851▼nKERIS▼z이 자료의 원문은 한국교육학술정보원에서 제공합니다.
