csci 599 digital geometry processing
play

CSCI 599: Digital Geometry Processing Spring 2015 Hao Li - PowerPoint PPT Presentation

CSCI 599: Digital Geometry Processing Spring 2015 Hao Li http://cs599.hao-li.com 1 USC Graphics http://gfx.usc.edu 2 Geometric Capture [Lab] http://www.hao-li.com 3 The Team Instructor ! Hao Li, hao.li@usc.edu Office: SAL 244


  1. CSCI 599: Digital Geometry Processing Spring 2015 Hao Li http://cs599.hao-li.com 1

  2. USC Graphics http://gfx.usc.edu 2

  3. Geometric Capture [Lab] http://www.hao-li.com 3

  4. The Team Instructor ! Hao Li, hao.li@usc.edu • Office: SAL 244 • Office hours: Tuesday 2-3PM • Assistants ! Kyle Olszewski, olszewsk@usc.edu • Pei-Lun Hsieh, peilun.hsieh@usc.edu • 4

  5. About Me

  6. Industrial Light & Magic

  7. Science, Engineering, & Art

  8. High Tech & Capital of Entertainment Disney DreamWorks Google Activision

  9. Introduction 9

  10. Target Audience ! • PhD students, MSc students, Advanced undergraduates • Computer Science , Computer Engineering, Mathematics, Physics, Game Program, Biomedicine, Bioengineering, etc. • Computer Graphics, Computer Vision, Robotics, Machine Learning, Signal and Image Processing, Medical Imaging ! 10

  11. Prerequisites ! • C/C++ Programming • Linear Algebra • Numerical Optimization ! • CSCI 420 Recommended 11

  12. Administrative When and where? ! Tuesday, Thursday, 11:00 am - 12:20 pm • KAP 158 (Kaprialian Hall) • Credits ! 3 Units • Website ! http://cs599.hao-li.com/ • 12

  13. Exercises Programming assignments ! based on OpenMesh (tutorial will be given Thursday next week) • cover some core stages of the geometry processing pipeline • C/C++ framework including 3D UI will be provided • Integral part of the lecture ! important for achieving course objectives • 13

  14. Grading Exercises Best 5 out of 6 exercises contribute to 70% of the final grade • Each exercise counts 20 points • Late submissions: every 5 minute removes 1 point in each exercise • Project Scope 2 months/person, Groups up to 2 • Implement a research paper around digital human capture but not • limited to it Final presentation, code/documentation, contributes 30% of the final • grade 14

  15. Academic Integrity • Do not copy any parts of the assignments from anyone • Do not look at other student’s code • Collaboration only for the project • USC Office of Student Judicial Affairs and Community Standards (Hell) will be notified 15

  16. Course Objectives • Define and relate the basic concept, tools, and algorithms in geometric modeling and digital geometry processing • Critically analyze and assess current research on surface representations and geometric modeling and apply the proposed methods in your own work • Design and implement individual components of geometric modeling system 16

  17. Recommended Textbook Botsch, Kobbelt, Pauly, Alliez, Levy: Polygon Mesh Processing , AK Peters, 2010 17

  18. Acknowledgement Course material taught at: ! EPFL, Mark Pauly (My PhD Advisor) • Bielefeld University, Mario Botsch • INRIA, Pierre Alliez, Bruno Levy • RWTH Aarchen, Leif Kobbelt • 18

  19. An Example 19

  20. Computer Graphics

  21. Performance Capture

  22. The Vision

  23. IMocap

  24. IMocap

  25. IMocap

  26. IMocap

  27. Facial Perfomance Capture 3 weeks for 10 seconds

  28. Geometry Capture Motion can be Captured at the Same Resolution as the Geometry

  29. Realtime Facial Performance Capture

  30. Capturing Geometry 30

  31. Static 3D Capture Stanford 2002

  32. Dynamic 3D Capture Stanford 2002

  33. Commercial 3D Capture Artec Group

  34. Full Body Capture 3D scanner 3D acquisition

  35. Multi-View Stereo Lee Perry-Smith, Infinite Realities + Agisoft

  36. Capturing Cities

  37. Google Earth

  38. Geometry ! γεω µ ετρία geo = earth metria = measure 38

  39. Geometry ! γεω µ ετρία 39

  40. ultrasound MRI scanner microscope Geometry ! x-ray diffractometer γεω µ ετρία stereo camera time-of-flight scanner radio telescope laser scanner 40

  41. Overview Geometric Modeling ! Techniques and algorithms for representing and processing • geometric objects We will focus on triangle meshes ! main questions: • why are triangles suitable representations for geometry • processing? what are the central processing algorithms? • how can they implemented efficiently? •

  42. Geometry Processing Pipeline 42

  43. Geometry Processing Pipeline 43

  44. Geometry Processing Pipeline 44

  45. Geometry Processing Pipeline 45

  46. Geometry Processing Pipeline 46

  47. Geometry Processing Pipeline 47

  48. Geometry Processing Pipeline 48

  49. Impacting Science

  50. Cardiology

  51. Evolutionary Biology

  52. Cancer Treatment

  53. Digitized Future 53

  54. For Everyone

  55. For Everyone

  56. For Everyone

  57. Scanning@Home 54°

  58. Living Room Entertainment

  59. In Tablet

  60. In Laptops

  61. In Laptops

  62. In Smartphones

  63. From Capture to Fabrication 3D printing 63

  64. Realtime Future 64

  65. Why Realtime? VFX/Game Production Virtual Avatars Robotics AR/Virtual Mirror

  66. Realtime Game Engines

  67. Realtime Facial Animation

  68. Virtual Reality Reloaded Oculus VR 2012 / Crytek 2014

  69. Personalized Future 69

  70. 3D Self-Portraits Omote3D Shashin Kan

  71. 3D Self-Portraits Omote3D Shashin Kan

  72. 3D Self-Portraits Omote3D Shashin Kan

  73. 3D Self-Portraits Omote3D Shashin Kan

  74. 3D Selfies

  75. 3D Selfies

  76. Personalized Games USC/ICT

  77. Personalized Applications MPI IS, Embodee entertainment fitness digital garment

  78. Fashion Industry

  79. Summary Geometry Processing Reconstruction Rendering Analysis Capture Reproduction Manipulation Design Simulation Storage 79

  80. Classic Graphics Geometry Processing Reconstruction Rendering Analysis Capture Reproduction Manipulation Design Simulation Storage 80

  81. Modern Graphics/Vision Geometry Processing Reconstruction Rendering Analysis Capture Reproduction Manipulation Design Simulation Storage 81

  82. The Future: Big Data / Robotics Geometry Processing Reconstruction Rendering Analysis Capture Reproduction Manipulation Design AI Simulation Feedback Storage 82

  83. Next Time • Parametric Approximations • Polygon Meshes • Data Structures 83

  84. http://cs599.hao-li.com Demos! 84

Recommend


More recommend