Honolulu, HI, U.S.A. | 21 July, 2017

Computer Vision and Pattern Recognition (CVPR 2017) Workshop

Deep Learning for Robotic Vision

Programme  Scope 

Live Stream

Afternoon Session Morning Session

Event Description

Recent advances in deep learning techniques have made impressive progress in many areas of computer vision, including classification, detection, and segmentation. While all of these areas are relevant to robotics applications, robotics also presents many unique challenges which require new approaches.
Challenges include the need for real-time analysis, the need for accurate 3d understanding of scenes, and the difficulty of doing experiments at scale. There are also opportunities which robotics brings to computer vision, for example, the ability to use depth sensors, to control where the camera is looking, and to provide a data source for "grounded" learning of concepts, reducing the need for manual labeling. We will consider work related to deep learning techniques in computer vision applied to a broad range of robotic devices, from self driving cars to drones to bipedal robots.

Invited Speakers

Jitendra Malik (UC Berkeley)
Raquel Urtasun (U Toronto / Uber ATG)
Dieter Fox (U Washington)
Honglak Lee (Google Brain / U Michigan)
Abhinav Gupta (CMU)
Jianxiong Xiao (AutoX)
Richard Newcombe (Facebook)
Raia Hadsell (Google DeepMind)
Ashutosh Saxena (Brain of Things)


Programme

08:00 - 08:15
Welcome and Introduction (The organisers)
08:15 - 08:45
Invited Talk: Ashutosh Saxena (Brain of Things)
08:45 - 09:15
Invited Talk: Richard Newcombe (Facebook)
09:15 - 09:45
Invited Talk: Jitendra Malik (UC Berkeley)
09:45 - 10:15
Poster Spotlights
10:15 - 10:45
Refreshment Break
10:45 - 12:00
Poster Session
12:00 - 13:30
Lunch Break
13:30 - 14:00
Invited Talk: Honglak Lee (Google Brain / U Michigan)
14:00 - 14:30
Invited Talk: Jianxiong Xiao (AutoX)
14:30 - 15:00
Invited Talk: Dieter Fox (U Washington)
15:00 - 15:30
Invited Talk: Raquel Urtasun (U Toronto / Uber ATG)
15:30 - 16:00
Refreshment Break
16:00 - 16:30
Invited Talk: Abhinav Gupta (CMU)
16:30 - 17:00
Invited Talk: Raia Hadsell (DeepMind)
17:00 - 17:55
Panel Discussion - Moderator: Kevin Murphy (Google Research), Juxi Leitner (ACRV/QUT), Niko Sünderhauf (ACRV/QUT)
17:55
Concluding Remarks

Poster Spotlights

Title Authors
Learning Robot Activities from First-person human Videos Using Convolutional Future Regression Jangwon Lee, Michael Ryoo (Indiana University)
End-to-End Driving in a Realistic Racing Game with Deep Reinforcement Learning Etienne Perot, Valeo; Maximilian Jaritz, Valeo/Inria; Marin Toromanoff, Valeo; Raoul de Charette (Inria)
Automated risk assessment for scene understanding and domestic robots using RGB-D data and 2.5D CNNs at a patch level Rob Dupre, Kingston University; Georgios Tzimiropoulos, Nottingham University; Vasileios Argyriou, Kingston University
Semantic Instance Segmentation for Autonomous Driving Bert De Brabandere, Davy Neven, Luc Van Gool, KU Leuven
Real-time hand grasp recognition using weakly supervised two-stage convolutional neural networks for understanding manipulation actions Ji Woong Kim, Sujeong You, Sang Hoon Ji, Hong Seok Kim, Korea Institute of Industrial Technology
Finding Anomalies with Generative Adversarial Networks for a Patrolbot Wallace Lawson, Esube Bekele, Keith Sullivan, Naval Research Laboratory
Time-Contrastive Networks: Self-Supervised Learning from Multi-View Observation Pierre Sermanet, Google; Corey Lynch, Google; Jasmine Hsu, Google; Sergey Levine, UC Berkeley
Curiosity-driven Exploration by Self-supervised Prediction Deepak Pathak, Pulkit Agrawal, Alyosha Efros, Trevor Darrell, UC Berkeley
 

Accepted Papers

Title Authors
3D Pose Regression using Convolutional Neural Networks Siddharth Mahendran, Haider Ali, Rene Vidal, Johns Hopkins University
Automated risk assessment for scene understanding and domestic robots using RGB-D data and 2.5D CNNs at a patch level Rob Dupre, Kingston University; Georgios Tzimiropoulos, Nottingham University; Vasileios Argyriou, Kingston University
Curiosity-driven Exploration by Self-supervised Prediction Deepak Pathak, Pulkit Agrawal, Alyosha Efros, Trevor Darrell, UC Berkeley
Detecting and Grouping Identical Objects for Region Proposal and Classification Wim Abbeloos, KU Leuven; Sergio Caccamo, KTH Royal Institute of Technology; Esra Ataer-Cansizoglu, Mitsubishi Electric Research Laboratories (MERL); Yuichi Taguchi, MERL; Chen Feng, MERL, Teng-Yok Lee, MERL
End-to-End Driving in a Realistic Racing Game with Deep Reinforcement Learning Etienne Perot, Valeo; Maximilian Jaritz, Valeo/Inria; Marin Toromanoff, Valeo; Raoul de Charette (Inria)
Episode-Based Active Learning with Bayesian Neural Networks Feras Dayoub, Niko Suenderhauf, Peter Corke, Queensland University of Technology
Finding Anomalies with Generative Adversarial Networks for a Patrolbot Wallace Lawson, Esube Bekele, Keith Sullivan, Naval Research Laboratory
Hand Movement Prediction Based Collision-free Human-Robot Interaction Yiwei Wang, Xin Ye, Yezhou Yang, Wenlong Zhang, Arizona State University
Learning Robot Activities from First-person human Videos Using Convolutional Future Regression Jangwon Lee, Michael Ryoo (Indiana University)
Leveraging Deep Reinforcement Learning for Reaching Robotic Tasks Kapil Katyal, I-Jeng Wang, Philippe Burlina, Johns Hopkins University
Real-time hand grasp recognition using weakly supervised two-stage convolutional neural networks for understanding manipulation actions Ji Woong Kim, Sujeong You, Sang Hoon Ji, Hong Seok Kim, Korea Institute of Industrial Technology
Semantic Instance Segmentation for Autonomous Driving Bert De Brabandere, Davy Neven, Luc Van Gool, KU Leuven
Time-Contrastive Networks: Self-Supervised Learning from Multi-View Observation Pierre Sermanet, Google; Corey Lynch, Google; Jasmine Hsu, Google; Sergey Levine, UC Berkeley
Tuning Modular Networks with Weighted Losses for Hand-Eye Coordination Fangyi Zhang, Jürgen Leitner, Michael Milford, Peter Corke, Queensland University of Technology

Call for Contributions

Topics

A PDF version of the Call for Contribution can be downloaded here!

We invite contributions (2 page extended abstracts) related to:
  • Deep learning for robotic vision.
  • Other computer vision techniques applied to robotics problems.
  • DNN based object recognition, detection and segmentation for robotics.
  • End-to-end perception algorithms.
  • Real-time algorithms for robotics perception.
  • Vision-based Simultaneous Localization and Mapping (SLAM).
  • 3D Scene understanding.
  • Deep learning in navigation and autonomous driving
  • Deep learning in human-robot interaction
  • Lifelong deep learning in robotics
  • Perception algorithms deployed on various robotic systems.
  • Reliable confidence measures for deep classifiers.
  • Deep learning for embedded systems and platforms with limited computational power
  • Deep learning for smart environments
  • Deep learning applications for the visually impaired and for the ageing society
  • Active perception.
  • Semi-supervised and self-supervised learning for robotics.



Abstract Submission Deadline

Submit via https://cmt3.research.microsoft.com/DLRV2017 by
April 7, 2017 (extended)
March 31, 2017

Author Notification

May 3, 2017
April 30, 2017

Camera Ready Submission

May 14, 2017

CVPR Workshop Date

July 21, 2017




Travel Grants

Travel grants are available for students or postdocs whose papers are accepted to the DLVR17 workshop. The grants will be awarded on a competitive basis to a number of applicants who have a need for financial support to participate in the workshop. The grants are intended to partially support travel, and will not cover all travel expenses.

If you would like to apply for such a grant, please e-mail to trung.pham@adelaide.edu.au prior to May 21, 2017 with the following information:
  1. Full name
  2. Paper title
  3. Institution (city, state, country)
  4. Degree and area of study
  5. Cover letter for your travel grant application (describing why you are applying).
You may also attach a resume.

Sponsors

Organisers

Primary contact: Anelia Angelova <anelia@google.com>

Anelia Angelova

Research Scientist

Google Research /
Google Brain

Gustavo Carneiro

Associate Professor

Australian Centre for Robotic Vision
University of Adelaide

Kevin Murphy

Research Scientist


Google Research


Niko Sünderhauf

Postdoctoral Fellow

Australian Centre for Robotic Vision
Queensland University of Technology

Ian Lenz

Postdoctoral Fellow


University of Texas, Austin

Vijay Kumar

Postdoctoral Fellow

Australian Centre for Robotic Vision
University of Adelaide

Postdoctoral Fellow

Australian Centre for Robotic Vision
Queensland University of Technology

Trung T. Pham

Postdoctoral Fellow

Australian Centre for Robotic Vision
University of Adelaide


Ingmar Posner

Lecturer


University of Oxford

Michael Milford

Associate Professor

Australian Centre for Robotic Vision
Queensland University of Technology


Wolfram Burgard

Professor


University of Freiburg

Ian Reid

Professor

Australian Centre for Robotic Vision
University of Adelaide

Peter Corke

Professor

Australian Centre for Robotic Vision
Queensland University of Technology


Program Committee

Kristen Grauman (University of Texas, Austin)
Oliver Brock (Technical University Berlin)
Yezhou Yang (Arizona State University)
Alex Kendall (Cambridge University)
Hema Koppula (Amazon)
Sergey Levine (University of California Berkeley)
Ashesh Jain (Stanford)
Paul Wohlhart (Google X)
George Papandreou (Google Research)
Stefan Leutenegger (Imperial College London)
Chelsea Finn (University of California Berkeley)
Fereshteh Sadeghi (University of Washington)
Nathan Silberman (Google Research)
Chenxia Wu (Cornell University)
Pierre Sermanet (Google Brain)
Kevin Lai (Amazon)
Edward Johns (Imperial College London)
David Held (Stanford)
Rodrigo Benenson (Max-Planck Institute for Informatics)