The first Workshop on
Structural and Compositional Learning on 3D Data


October 16 @ ICCV 2021 Virtual


  • Home

Select your time zone
Default: Eastern Daylight Time

Start End Event

Welcome and Introductions

[Slides PDF Link]
Workshop Organizers (Kaichun Mo)
Session 1: Structural Representations for 3D Scenes

Opening Remarks for Session 1

Workshop Organizers (Kai Wang, Antoni Rosinol)

[Spotlight] Describing 3D Indoor Scenes via Semantic Scene Graphs


Helisa Dhamo (TUM)

[Spotlight] Compositional Structure in 3D Vision and Language


Siyuan Huang (UCLA)

[Keynote] Learning to Infer and Generate Programs for 3D Shapes and Scenes


Daniel Ritchie (Brown)

[Keynote] 3D Dynamic Scene Graphs: High-level Scene Understanding for Embodied Intelligence


Luca Carlone (MIT)
Session 2: Compositional Structures in Robotics

Opening Remarks for Session 2

Workshop Organizers (Fei Xia, Danfei Xu)

[Keynote] Structure from Action: Object Structure Discovery with Active Interactions


Shuran Song (Columbia University)

[Spotlight] Taskography - Robot task planning over large 3D scene graphs


Krishna Murthy (Mila)

[Spotlight] Hierarchical Planning for Long-Horizon Manipulation with Geometric and Symbolic Scene Graphs


Yifeng Zhu (UT Austin)

[Keynote] Learning Visual Representations, Language, and Scene Dynamics via Interaction with the World


Roozbeh Mottaghi (AI2 & University of Washington)

Accepted Paper Poster Session 1
Session 3: Structural Representations for 3D Shapes

Opening Remarks for Session 3

Workshop Organizers (Minhyuk Sung, Kaichun Mo)

[Keynote] 3D Structure Extraction and Generation


Leonidas J. Guibas (Stanford)

[Keynote] Unsupervised Learning of 3D Shape Structures


Hao (Richard) Zhang (SFU)

[Spotlight] ParSeNet: A Parametric Surface Fitting Network for 3D Point Clouds


Gopal Sharma (UMass Amherst)

[Spotlight] Composing Humans and Objects in the Wild


Jason Zhang (CMU)
Session 4: 3D Structural and Compositional Discovery

Opening Remarks for Session 4

Workshop Organizers (Paul Guerrero, Songfang Han)

[Keynote] Hybrid explicit/implicit representations of shape


Thomas Funkhouser (Princeton & Google)

[Keynote] Compositional Learning in Geometry


Hao Su (UCSD)

[Spotlight] Local Deep Implicit Functions


Kyle Genova (Google Research)

[Spotlight] Neural Parts: Learning Expressive 3D Shape Abstractions with Invertible Neural Networks


Despoina Paschalidou (Max Planck ETH)

Panel Discussion

Host: Shubham Tulsiani, Niloy Mitra
Panelists: Shuran Song, Roozbeh Mottaghi, Hao Su, Hao (Richard) Zhang, Daniel Ritchie, Thomas Funkhouser.

Accepted Paper Poster Session 2

Contact Info

E-mail: struco3d@googlegroups.com or kaichun@cs.stanford.edu

Acknowledgements

Website template borrowed from: https://futurecv.github.io/ (Thanks to Deepak Pathak)