3D vision is currently undergoing a period of tremendous growth. For instance, there has been an explosion of 3D big data: depth sensors like Kinect now sit in the living rooms of millions of consumers and repositories like 3D Warehouse and Shapenet provide the millions of instances required by large-scale learning systems. Driven by this, a wide variety of classic 3D vision problems, ranging from 3D estimation from still images to 3D model retrieval have seen dramatic performance gains, enabling fundamentally new capabilities.

Despite the tremendous progress in 3D vision, there has been less effort devoted to applying this hard-won progress to other disciplines of machine intelligence: for instance, even though the world is fundamentally 3D, semantic recognition and captioning tasks are still typically solved without any consideration of 3D. Similarly, 3D vision does not take full advantage of work in areas such as graphics or robotics.

The theme of this workshop is how we can bridge the gap between 3D vision and other fields and examine how 3D vision can help and be helped by other areas of machine intelligence. Our goal is to highlight work that tries to join 3D vision and other disciplines, bring together experts in multiple domains, and to encourage discussion of how to connect with 3D vision. We are particularly interested in work that joins 3D vision with robotics, graphics, natural language processing, virtual reality, and semantics.

Invited Speakers


Derek Hoiem

Vladlen Koltun

Jitendra Malik

University of Illinois at Urbana-Champaign
Intel Visual Computing Lab
University of California, Berkeley

Noah Snavely

Raquel Urtasun

Cornell University
University of Toronto

Date and Time

Tentative Schedule

1:50-2:00Welcome and Overview
2:00-2:30 Noah Snavely
2:30-3:00 Derek Hoiem
3:00-3:30 Vladlen Koltun
3:30-4:30Poster Session and Coffee Break
4:30-5:00 Raquel Urtasun
5:00-5:30 Jitendra Malik
5:30-6:00Panel Discussion


We've selected a mix of papers from CVPR 2017 as well as past work from other non-CVPR/ICCV/ECCV conferences. Our aim is to get people with diverse perspective on 3D in the same room and talking about both problems and tools for solving them.

Generating Holistic 3D Scene Abstractions for Text-Based Image Retrieval
Ang Li, Jin Sun, Joe Yue-Hei Ng, Ruichi Yu, Vlad I. Morariu, Larry S. Davis
CVPR 2017

Cross-Modal Adaptation for RGB-D Detection
Judy Hoffman, Saurabh Gupta, Jian Leong, Sergio Guadarrama, Trevor Darrell
ICRA 2016

Body Talk: Crowdshaping Realistic 3D Avatars with Words
Stephan Streuber, M. Alejandra Quiros-Ramirez, Matthew Q. Hill, Carina A. Hahn, Silvia Zuffi, Alice O’Toole, Michael J. Black

Hamid Izadinia, Qi Shan, Steven M. Seitz
CVPR 2017

Learning from Synthetic Humans
Gul Varol, Javier Romero, Xavier Martin, Naureen Mahmood, Michael J. Black, Ivan Laptev, Cordelia Schmid

Semantic Multi-View Stereo: Jointly Estimating Objects and Voxels
Ali Osman Ulusoy, Michael J. Black, Andreas Geiger
CVPR 2017

Deep Supervision with Shape Concepts for Occlusion-Aware 3D Object Parsing
Chi Li, M. Zeeshan Zia, Quoc-Huy Tran, Xiang Yu, Gregory D. Hager, Manmohan Chandraker
CVPR 2017

Learning to Align Semantic Segmentation and 2.5D Maps for Geolocalization
Anil Armagan, Martin Hirzer, Peter M. Roth, Vincent Lepetit
CVPR 2017

Learning a Probabilistic Latent Space of Object Shapes via 3D Generative-Adversarial Modeling
Jiajun Wu, Chengkai Zhang, Tianfan Xue, William T. Freeman, Joshua B. Tenenbaum
NIPS 2016

Visual-Inertial-Semantic Scene Representation for 3D Object Detection
Jingming Dong, Xiaohan Fei, Stefano Soatto
CVPR 2017

Text to 3D Scene Generation with Rich Lexical Grounding
Angel X. Chang, Will Monroe, Manolis Savva, Christopher Potts, Christopher D. Manning
Proceedings of ACL 2015

PiGraphs: Learning Interaction Snapshots from Observations
Manolis Savva, Angel X. Chang, Pat Hanrahan, Matthew Fisher, Matthias Nießner

Transformation-Grounded Image Generation Network for Novel 3D View Synthesis
Eunbyung Park, Jimei Yang, Ersin Yumer, Duygu Ceylan, Alexander C. Berg
CVPR 2017



David Fouhey

Qixing Huang

Joseph Lim

University of California, Berkeley
University of Texas at Austin
University of Southern California

Contact Info

E-mail: dfouhey at eecs dot berkeley dot edu