SIGGRAPH 2022: Neural Geometry Processing

SIGGRAPH 2022勉強会で使ったスライドです。
https://siggraph.xyz/s2022/

Transcript

1. Neural Geometry Processing Tatsuya Yatagawa Aug. 27th, 2022 SIGGRAPH 2022 Paper Reading Seminar
2. 6 TOG , 2 Conference Track l Dual Octree Graph Networks for Learning Adaptive Volumetric ... l DiffusionNet: Discretization Agnostic Learning on Surfaces l Neural Dual Contouring l Random Walks for Adversarial Meshes l PCEDNet: A Lightweight Neural Network for Fast and Interactive ... l Self-Sampling for Neural Point Cloud Consolidation l ImLoveNet: Misalignment Image-supported Registration Network ... l DSG-Net: Learning Disentangled Structure and Geometry ... Neural Geometry Processing (8 papers)
3. Dual Octree Graph Networks for Learning Adaptive Volumetric Shape Representations CNN ,
4. 1. 2. , 3. MLP Dual Octree Graph
5. l , . l .
6. DiffusionNet: Discretization Agnostic Learning on Surfaces
7. : l 𝑡 , DiffusionNet , (= ) l , : , : +
8. , l target l , , . : , attention . , .
9. Neural Dual Contouring Dual Contouring
10. Marching Cubes ( : Dual Contouring, : Marching Cubes) Dual Contouring Dual Contouring , Marching Cubes
11. l MC (MC ) l MC , , : Neural Marching Cubes l , (= ) l
12. l NMC , DC l NMC , DC l NDC , UNDC l DC l UNDC , Neural Dual Contouring
13. l ( ) , ,
14. Random Walks for Adversarial Meshes (Conf. Track) ,
15. , (adversarial sample) : [Goodfellow et al., ICLR 2015] l
16. l RandomWalker [Lahav and Tal 2020] l : Imitating Network
17. l , random walk l Imitating Net , :
18. l , (RandomWalker, MeshNet) l Imitating Network (Walk path Random Walker )
19. PCEDNet: A Lightweight Neural Network for Fast and Interactive Edge Detection in 3D Point Clouds
20. (Scale-Space Matrix) l Growing Least Squares [Mellado+ 2012] l , - , , , . l 6 , 16 (=SSM) . PCEDNet (Point Cloud Edge Detection Network) l SSM (= 6 16 ) .
21. l , , , l MLP, CNN l CNN , , l SSM 4-128 , l , 16 4 16 128
22. Self-Sampling for Neural Point Cloud Consolidation , Input Consolidated Input Consolidated
23. : l positive ( ) negative ( ) . l , (source) (target) , positive l
24. , . Input Consolidated Input Consolidated : , positive, negative , .
25. ImLoveNet: Misaligned Image-supported Registration Network for Low-overlap Point Cloud Pairs (Conf. Track) 2
26. : , 2 l (DGR [Choy+ 2020], PREDATOR [Huang+ 2021] etc.). l )
27. SOTA (PREDATOR, OMNet) : , , .
28. DSG-Net: Learning Disentangled Structure and Geometry for 3D Shape Generation (VAE)
29. : , . : .
30. l 2 + ( ) ( ) l as-consistent-as- possible (9 BBox ) [Guo+ 2019] l one-hot MLP l ACAP BBox , .
31. (SDM-Net) , ,
32. 6 TOG , 2 Conference Track l Dual Octree Graph Networks for Learning Adaptive Volumetric ... l DiffusionNet: Discretization Agnostic Learning on Surfaces l Neural Dual Contouring l Random Walks for Adversarial Meshes l PCEDNet: A Lightweight Neural Network for Fast and Interactive ... l Self-Sampling for Neural Point Cloud Consolidation l ImLoveNet: Misalignment Image-supported Registration Network ... l DSG-Net: Learning Disentangled Structure and Geometry ... : Neural Geometry Processing : , . , (= ) .