Recovering 6D Object Pose Estimation


The development of RGB-D sensors, high GPU computing, and scalable machine learning algorithms have opened the door to a whole new range of technologies and applications which require detecting and estimating object poses in 3D environments for a variety of scenarios.Our program will feature several high-quality invited talks, poster presentations, and a panel discussion to identify key research questions and highlight future research directions. We invite submissions of papers and extended abstracts of ongoing or already published work. Accepted papers and extended abstracts will be presented as orals/posters/spotlights at the workshop, in areas including, but not limited to:
  • 3D object detection
  • Object detection in depth images
  • Joint registration of multiple 3D objects
  • 6 DoF object pose, under heavy occlusions and background clutters
  • Occlusion-aware segmentation
  • Bin-picking
  • 3D object tracking
  • Surface representation and fitting
  • Non-rigid (deformable or articulated) object registration
  • 3D object modelling
  • Multiple object instance detection
  • Object class (category) detection

We also invite authors to submit results to our challenges. Such results will be presented at the workshop and likely cited in later work.

Previous R6D Workshop: R6D2015

Important Dates

Paper Submission Deadline: 26 Aug

Notification of Acceptance: 2 Sep

Camera Ready: 9 Sep

Extended Abstract Submission Deadline: 2 Sep

Challenge Submission Deadline: 9 Sep

Workshop Date: 9 Oct

Venue

R6D Workshop is part of the ECCV 2016 workshops.

Please see the ECCV webpage for details on venue, accomodations etc.

News

Our website is online

Participate

Papers / Extended Abstracts Submissions

We invite submissions of new papers and extended abstracts of already published or ongoing work. Authors are given the opportunity to promote their work to an extended audience in all areas covered by the workshop. Accepted papers and abstracts will be presented as orals/posters/spotlights at the workshop.

Extended abstracts are limited to two pages. Authors are free to use any format available, e.g. the ECCV 2016 word or LaTex templates tar or zip. Extended abstracts should be sent directly to r6d.workshop@gmail.com

Paper submissions will be handled electronically via the conference’s CMT Website.

Challenges

Following the success and popularity of other challenges in object detection in 2D images, we want to establish a new challenge (comprised of 4 sub-challenges) in the area of 6 DoF object detection, which, to our knowledge, is the first of each kind. Each challenge will include and gradually be enriched with numerous test scenes and objects, including their 3D models. You can apply your method to our data and submit your results. We will evaluate submitted results according to multiple metrics and display the scores for comparison. For more information please visit the respective website of the individual challenge.

Imperial College London   TU Dresden (articulated)   TU Dresden (occluded)   TU Graz & EPFL   Czech Tech. University   University of Birmingham

Awards

Best Paper Award

A Direct Method for Robust Model-Based 3D Object Tracking from a Single RGB Image
Byung-Kuk Seo, Wuest Harald

Best Poster Award

Deep Learning of Local RGB-D Patches for 3D Object Detection and 6D Pose Estimation
Wadim Kehl, Fausto Milletari, Federico Tombari, Slobodan Ilic, Nassir Navab

Keynote Speakers

Program

08:50 - 09:00 Opening
09:00 - 09:30 Invited talk: Silvio Savarese
09:30 - 10:00 Invited talk: Carsten Steger
10:00 - 10:30 Coffee break
10:30 - 11:30 Oral Session
Paper IDTitle
1Direct-from-Video: Unsupervised NRSfM, Karel Lebeda, Simon Hadfield, Richard Bowden
2Towards Categorization and Pose Estimation of Sets of Occluded Objects in Cluttered Scenes from Depth Data and Generic Object Models Using Joint Parsing, Hector Basevi, Ales Leonardis
3Reconstructing Articulated Rigged Models from RGB-D Videos, Dimitrios Tzionas, Juergen Gall
4On Evaluation of 6D Object Pose Estimation, Tomas Hodan, Jiri Matas, Stepan Obdrzalek
11:30 - 12:00 Invited talk: Jianxiong Xiao
12:00 - 13:30 Lunch break
13:30 - 14:00 Spotlights
14:00 - 15:00 Poster Session
Poster IDTitle
1RobotFusion: Grasping with a Robotic Manipulator via Multi-view Reconstruction, Daniele De Gregorio, Federico Tombari, Luigi Stefano
2A Direct Method for Robust Model-Based 3D Object Tracking from a Single RGB Image, Byung-Kuk Seo, Wuest Harald
3A Radial Search Method for fast Nearest Neighbor Search on Range Images, Federico Tombari, Samuele Salti, Luca Puglia, Luigi Stefano, Giancarlo Raiconi
4Physical Reasoning for 3D Object Recognition using Global Hypothesis Verification, Shuichi Akizuki, Manabu Hashimoto
5SemanticFusion: Joint Labeling, Tracking and Mapping, Tommaso Cavallari, Luigi Stefano
6Affordance-focused Features for Generic Object Recognition, Shuichi Akizuki, Masaki Iizuka, Manabu Hashimoto
7Fast Feature-Less Quaternion-based Particle Swarm Optimization for Rigid and Articulated Pose Estimation From RGB-D Images, Giorgio Toscana, Stefano Rosa
8Object Detection for Random Bin Picking using Point Pair Features, Wim Abbeloos, Toon Goedeme
9Fast 6D Pose Estimation from a Monocular Image using Hierarchical Pose Trees, Yoshinori Konishi, Yuki Hanzawa, Masato Kawade, Manabu Hashimoto
10ObjectNet3D: A Large Scale Database for 3D Object Recognition, Yu Xiang, Wonhui Kim, Wei Chen, Jingwei Ji, Christopher Choy, Hao Su, Roozbeh Mottaghi, Leonidas J. Guibas, Silvio Savarese
11SDF-2-SDF: Highly Accurate 3D Object Reconstruction, Miroslava Slavcheva, Wadim Kehl, Nassir Navab, Slobodan Ilic
12Bayesian Image based 3D Pose Estimation, Valsamis Ntouskos, Fiora Pirri, Marta Sanzari
13Learning a Predictable and Generative Vector Representation for Objects, Rohit Girdhar, David Fouhey, Mikel Rodriguez, Abhinav Gupta
14Localizing and Orienting Street Views Using Overhead Imagery (location and orientation of street images), Nam Vo, James Hays
15Weakly Supervised Object Localization Using Size Estimates, Miaojing SHI, Vittorio Ferrari
16Real-time Joint Tracking of a Hand Manipulating an Object from RGB-D Input, Srinath Sridhar, Franziska Mueller, Michael Zollhoefer, Dan Casas, Antti Oulasvirta, Christian Theobalt
17Global Registration of 3D Point Sets via LRS decomposition, Federica Arrigoni, Beatrice Rossi, Andrea Fusiello
184D Match Trees for Non-rigid Surface Alignment, Armin Mustafa, Hansung Kim, Adrian Hilton
19Non-Rigid 3D Shape Retrieval via Large Margin Nearest Neighbor Embedding, Ioannis Chiotellis, Rudolph Triebel, Thomas Windheuser, Daniel Cremers
20Deep learning 3D shape surfaces using geometry images, Ayan Sinha, Jing Bai, Karthik Ramani
21Single Image 3D Interpreter Network, Jiajun Wu, Tianfan Xue, Joseph Lim, Yuandong Tian, Joshua Tenenbaum, Antonio Torralba, William Freeman
223D-R2N2: A unified approach for single and multi-view 3D object reconstruction, Christopher Choy, Danfei Xu, JunYoung Gwak, Kevin Chen, Silvio Savarese
23Deep Deformation Network for Object Landmark Localization, Xiang Yu, Feng Zhou, Manmohan Chandraker
24DOC: Deep OCclusion Estimation From a Single Image, Peng Wang, Alan Yuille
25Facilitating and Exploring Planar Homogeneous Texture for Indoor Scene Understanding, Shahzor Ahmad, Loong-Fah Cheong
26Hand pose estimation from local surface normals, Chengde Wan, Angela Yao, Luc Van Gool
27Deep Learning of Local RGB-D Patches for 3D Object Detection and 6D Pose Estimation, Wadim Kehl, Fausto Milletari, Federico Tombari, Slobodan Ilic, Nassir Navab
28Multi-view 3D Models from Single Images With a Convolutional Network, Maxim Tatarchenko, Alexey Dosovitskiy, Thomas Brox
15:00 - 15:30 Coffee break
15:30 - 16:00 Invited talk: Tinne Tuytelaars
16:00 - 16:30 Invited talk: Zoltan Marton
16:30 - 17:00 Invited talk: Carsten Rother
17:00 - 17:20 Awards and closing

Workshop Chairs

Tae-Kyun Kim, Imperial College London

Jiri Matas, Chech Technical University

Vincent Lepetit, Technical University Graz

Carsten Rother, Technical University Dresden

Ales Leonardis, University of Birmingham

Krzysztof Wallas, Poznan University of Technology

Carsten Steger, MVTec GmbH

Rigas Kouskouridas, Imperial College London

Challenge Chairs

Frank Michel, Technical University Dresden

Alexander Krull, Technical University Dresden

Andreas Doumanoglou, Imperial College London

Tomas Hodan, Chech Technical University

Krzysztof Wallas, Poznan University of Technology

Alberto Crivellaro, EPFL

Technical Program Committee

Antonis Argyros, FORTH

Ujwal Bonde, Toshiba Research

Eric Brachman, Technical University Dresden

Alessio Del Bue, IIT

Bertram Drost, MVTec GmbH

Ozgur Erkent, University of Innsbruck

Hironobu Fujiyoshi, Chubu University

Stefan Hinterstoisser, Google

Slobodan Ilic, Siemens

Edward Johns, Imperial College London

Hamidreza Kasaei, University of Porto

Wadim Kehl, Technical University Munich

Yoshinori Konishi, Omron

Victor Lempitsky, Skoltech

Joseph J. Lim, Stanford

Justus Piater, University of Innsbruck

Colin Rennie, Rutgers University

Caner Sahin, Imperial College London

David Tan, Technical University Munich

Damien Teney, University of Adelaide

Ulrike Thomas, Technical University of Chemnitz