报告论文安排

时间：2022年01月07日（周五） 13:00-17:30 19:00-21:30

地点：未来中心3号报告厅 A321

0. 报告属于正常上课，所有同学不得无故缺席。
1. 每篇论文报告总时长为12分钟左右，其中陈述不超过10分钟（包括展示video），提问2分钟左右，建议ppt不超过15页！
2. 由于每篇论文的报告时间有限，比较短，因此报告的重点在于：简单介绍作者和单位情况； 主要讲一下这篇文章在做什么(What)？为什么做这个问题(motivation-Why)？这个问题以前有些什么工作？ 难点在哪里？这篇文章解决了什么样的难点或提出什么样的新方法(selling points)？局限和未来工作(limitation and future work)？ 细节若太多不用展开，关键要让大家知道这篇文章的主要思想和价值所在。
3. 请大家尽量在网上去找关于你所负责报告论文的相关资料。如各个作者的主页，项目网页。 很多作者会把论文、video、甚至将PPT放在他的主页。讲解论文时，论文中的图片、视频均播放给大家看一下。 “一图胜千言”，视频又能胜于图片。
4. 书面报告要求采用科研论文格式（标题、摘要、关键字、正文、参考文献等）撰写， 内容可以是所报告论文的概览、同主题多篇论文的综述等。
5. 评分：PPT口头报告（小组） + 书面报告（个人）

SIGGRAPH 2021 papers: 60

TOG: Vol 40, No 4 (acm.org)

SIGGRAPH 2021 Papers (realtimerendering.com)

1. Hierarchical neural reconstruction for path guiding using hybrid path and photon samples
2. Real-time neural radiance caching for path tracing
3. Interactive Monte Carlo denoising using affinity of neural features
4. Weakly-supervised contrastive learning in path manifold for Monte Carlo image reconstruction
5. Total relighting: learning to relight portraits for background replacement
6. PhotoApp: photorealistic appearance editing of head portraits
7. Only a matter of style: age transformation using a style-based regression model
8. Coarse-to-fine: facial structure editing of portrait images via latent space classifications
9. Fusion 360 gallery: a dataset and environment for programmatic CAD construction from human design sequences
10. Neural Light Transport for Relighting and View Synthesis
11. ROSEFusion: random optimization for online dense reconstruction under fast camera motion
12. Neural complex luminaires: representation and rendering
13. Acorn: adaptive coordinate networks for neural scene representation
14. Mixture of volumetric primitives for efficient neural rendering
15. Scalable image-based indoor scene rendering with reflections
16. DeepFormableTag: end-to-end generation and recognition of deformable fiducial markers
17. High-order differentiable autoencoder for nonlinear model reduction
18. End-to-end complex lens design with differentiate ray tracing
19. Neural monocular 3D human motion capture with physical awareness
20. MoCap-solver: a neural solver for optical motion capture data
21. Capturing detailed deformations of moving human bodies
22. TransPose: real-time 3D human translation and pose estimation with six inertial sensors
23. Real-time 3D neural facial animation from binocular video
24. Learning an animatable detailed 3D face model from in-the-wild images
25. Deep relightable appearance models for animatable faces
26. DeepFaceEditing: deep face generation and editing with disentangled geometry and appearance control
27. Neural animation layering for synthesizing martial arts movements
28. Learning a family of motor skills from a single motion clip
29. Real-time deep dynamic characters
30. Seamless manga inpainting with semantics awareness
31. Learning meaningful controls for fluids
32. TryOnGAN: body-aware try-on via layered interpolation
33. StyleCariGAN: caricature generation via StyleGAN feature map modulation
34. AgileGAN: stylizing portraits by inversion-consistent transfer learning
35. ManipNet: neural manipulation synthesis with a hand-object spatial representation
36. Highlight-aware two-stream network for single-image SVBRDF acquisition
37. Free-form scanning of non-planar appearance with neural trace photography
38. The effect of shape and illumination on material perception: model and applications
39. Learning active quasistatic physics-based models from data
40. Learning skeletal articulations with neural blend shapes
41. Learning contact corrections for handle-based subspace dynamics
42. Designing an encoder for StyleGAN image manipulation
43. SWAGAN: a style-based wavelet-driven generative model
44. Driving-signal aware full-body avatars
45. Learning time-critical responses for interactive character control
46. Consistent depth of moving objects in video
47. Editable free-viewpoint video using a layered neural representation
48. Video recoloring via spatial-temporal geometric palettes
49. SP-GAN: sphere-guided 3D shape generation and manipulation
50. Unsupervised learning for cuboid shape abstraction via joint segmentation from point clouds
51. ShapeMOD: macro operation discovery for 3D shape programs
52. Volumetric appearance stylization with stylizing kernel prediction network
53. Neural scene graph rendering
54. Orienting point clouds with dipole propagation
55. HodgeNet: learning spectral geometry on triangle meshes
56. Eliminating topological errors in neural network rotation estimation using self-selecting ensembles
57. NeuMIP: multi-resolution neural materials
58. Optimizing UI layouts for deformable face-rig manipulation
59. A gradient-based framework for 3D print appearance optimization
60. Displaced signed distance fields for additive manufacturing

SIGGRAPH Asia 2021 papers: 38

TOG: Vol 40, No 6 (acm.org)

SIGGRAPH Asia 2021 Papers (realtimerendering.com)

61. Transflower: probabilistic autoregressive dance generation with multimodal attention
62. Motion recommendation for online character control
63. SuperTrack: motion tracking for physically simulated characters using supervised learning
64. PBNS: physically based neural simulation for unsupervised garment pose space deformation
65. Modeling clothing as a separate layer for an animatable human avatar
66. Predicting high-resolution turbulence details in space and time
67. Human dynamics from monocular video with dynamic camera movements
68. Camera keyframing with style and control
69. Layered neural atlases for consistent video editing
70. Aesthetic-guided outward image cropping
71. Time-travel rephotography
72. Barbershop: GAN-based image compositing using segmentation masks
73. SketchHairSalon: deep sketch-based hair image synthesis
74. EyelashNet: a dataset and a baseline method for eyelash matting
75. Pose with style: detail-preserving pose-guided image synthesis with conditional StyleGAN
76. Neural actor: neural free-view synthesis of human actors with pose control
77. Live speech portraits: real-time photorealistic talking-head animation
78. Semi-supervised video-driven facial animation transfer for production
79. Rendering with style: combining traditional and neural approaches for high-quality face rendering
80. FreeStyleGAN: free-view editable portrait rendering with the camera manifold
81. Aerial path planning for online real-time exploration and offline high-quality reconstruction of large-scale urban scenes
82. AutoMate: a dataset and learning approach for automatic mating of CAD assemblies
83. Dynamic neural garments
84. Neural radiosity
85. NeRFactor: neural factorization of shape and reflectance under an unknown illumination
86. HyperNeRF: a higher-dimensional representation for topologically varying neural radiance fields
87. Neural frame interpolation for rendered content
88. Neural 3D holography: learning accurate wave propagation models for 3D holographic virtual and augmented reality displays
89. Binaural audio generation via multi-task learning
90. Intuitive and efficient roof modeling for reconstruction and synthesis
91. Deep3DLayout: 3D reconstruction of an indoor layout from a spherical panoramic image
92. Neural marching cubes
93. TM-NET: deep generative networks for textured meshes
94. I♥LA: compilable markdown for linear algebra
95. DeepVecFont: synthesizing high-quality vector fonts via dual-modality learning
96. Monte Carlo denoising via auxiliary feature guided self-attention
97. Learning to cluster for rendering with many lights
98. Generative modelling of BRDF textures from flash images