Research Group on Visual Computation

• Home
• Projects
• Open positions
• ICCV 2013 Summary

• Zoltan Kato

ICCV 2013 Summary

(Author: Zoltan Kato)

ICCV is the premier international Computer Vision event comprising the main ICCV conference and several co-located workshops and short courses. In 2013, ICCV took place at the Sydney Conference Centre in Darling Harbour, Sydney, December 1-8.

Herein, I give a brief summary on the publications/lectures relevant to our reserach areas. Most of the papers are publicly available at the Computer Vision Papers website.

Tutorials

Roland Angst: Tutorial on Low-Dimensional Subspaces in Computer Vision

Abstract:
Being a highly general concept, subspaces represent basic buildings blocks for more complex models and abstractions, and are therefore prevalent in science and engineering. There exists a wide variety of subspace models and algorithms, usually developed independently by different research communities. Whereas principal component analysis and statistical factor models represent very general modeling tools, more specific subspace-based models are given for example by recommender systems or multiple view tensors. Due to their practical importance, subspaces and basic concepts building upon them are taught early on in any engineering or science curriculum. However, more recent and complex models involving subspaces enjoy a rich geometric and algebraic structure which clearly go beyond these basic concepts. Moreover, subspaces are inherently linked to rank- or dimensionality-constraints which can quickly render formulations into highly non-trivial optimization problems.
More details

Main Conference

3D Reconstruction, Stereo

Rolling Shutter Stereo, Olivier Saurer, ETH Zurich; Kevin Koeser; Jean-Yves Bouguet, Google; Marc Pollefeys, ETH
Elastic Fragments for Dense Scene Reconstruction, Qian-Yi Zhou, Stanford University; Stephen Miller, Stanford University; Vladlen Koltun, Stanford University
A Rotational Stereo Model Based On XSlit Imaging, Jinwei Ye, University of Delaware; Yu Ji, University of Delaware; Jingyi Yu, University of Delaware
Lifting 3D Manhattan Lines from a Single Image, Srikumar Ramalingam, MERL; Matthew Brand, MERL
Point-Based 3D Reconstruction of Thin Objects, Benjamin Ummenhofer, University of Freiburg; Thomas Brox
Multi-View 3D Reconstruction from Uncalibrated Radially-Symmetric Cameras, Jae-Hak Kim, Australian National University; Yuchao Dai, Australian National University; Hongdong Li, Australian National University and NICTA; Xin Du, Zhejiang University; Jonghyuk Kim, Australian National University. (Code is available from the authors!)
A Unified Rolling Shutter and Motion Blur Model for Dense 3D Visual Tracking, Maxime Meilland, I3S-CNRS-UNS; Tom Drummond, Monash University; Andrew Comport, CNRS-I3S/UNS
A Flexible Scene Representation for 3D Reconstruction Using an RGB-D Camera, Diego Thomas, National Institute of Informatics; Akihiro Sugimoto

Correspondence Problem, Repetitive Structures

Network Principles for SfM: Disambiguating Repeated Structures with Local Context, Kyle Wilson, Cornell University; Noah Snavely, Cornell University
PM-Huber: PatchMatch with Huber Regularization for Stereo Matching, Philipp Heise, TU Mnchen; Sebastian Klose, TU Mnchen; Brian Jensen, TU Mnchen; Alois Knoll, TU Mnchen
Robust Non-parametric Data Fitting for Correspondence Modeling, Wen-Yan Lin, Oxford Brookes; Ming-Ming Chen; Shuai Zheng; Jiangbo Lu, Advanced Digital Sciences Cent; Nigel Crook

Camera Calibration and Pose Estimation

A Global Linear Method for Camera Pose Registration, Nianjuan Jiang, ADSC; Zhaopeng Cui, NUS; Ping Tan
Pose Estimation with Unknown Focal Length using Points, Directions and Lines, Yubin Kuang, Lund University; Kalle Aström, Lund University
Unsupervised Intrinsic Calibration from a Single Frame Using a Plumb-Line Approach, R. Melo, ISR-Coimbra; M. Antunes; J.P. Barreto; G. Falcao; N. Gonalves
An Enhanced Structure-from-Motion Paradigm Based on the Absolute Dual Quadric and Images of Circular Points, Lilian Calvet, University of Toulouse; Pierre Gurdjos, IRIT
Camera Alignment Using Trajectory Intersections in Unsynchronized Videos, Thomas Kuo, UC Santa Barbara; Santhoshkumar Sunderrajan, UCSB; B.S. Manjunath, UCSB
Revisiting the PnP Problem: A Fast, General and Optimal Solution, Yinqiang Zheng, Tokyo Institute of Technology; Yubin Kuang, Lund University; Shigeki Sugimoto, Tokyo Institute of Technology; Kalle Aström, Lund University; Masatoshi Okutomi, Tokyo Institute of Technology
Direct Optimization of Frame-to-Frame Rotation, Laurent Kneip, ETH Zurich; Simon Lynen, ETH Zurich
Extrinsic Camera Calibration without A Direct View Using Spherical Mirror, Amit Agrawal, MERL
Global Fusion of Relative Motions for Robust, Accurate and Scalable Structure from Motion, Pierre Moulon, Université Paris-Est, LIGM (UMR CNRS), ENPC, Mikros Image; Pascal Monasse, Université Paris-Est, LIGM (UMR CNRS), ENPC ; Renaud Marlet, Université Paris-Est, LIGM (UMR CNRS), ENPC