Eleventh IEEE International Conference on Computer Vision
Rio de Janeiro, Brazil, October 14-20, 2007
Ioannis Kakadiaris (firstname.lastname@example.org)
Nikos Paragios (email@example.com)
Oct. 14th, Morning
Optical Motion Capture
Lecturers: Yiannis Aloimonos, Gutemberg Guerra-Filho
We discuss the theoretical and empirical aspects of optical motion capture systems. The problems involved are initialization, marker/feature detection, spatial correspondence (stereo), temporal correspondence (tracking), and post-processing. We present the theory involved in each problem and techniques used in its implementation.
Human-Centered Vision Systems
Lecturers: Thomas S. Huang, Alejandro (Alex) Jaimes, Nicu Sebe
This tutorial takes a holistic view on the research issues and applications of Human-Centered Vision Systems focusing on three main areas: (1) multimodal interaction: visual (body, gaze, gesture) and audio (emotion) analysis; (2) image databases, indexing, and retrieval: context modeling, cultural issues, and machine learning for user-centric approaches; (3) multimedia data: conceptual analysis at different levels (feature, cognitive, and affective).
Oct. 14th, Afternoon
Gradient Domain Manipulation Techniques in Vision and Graphics
Lecturers: Amit Agrawal, Ramesh Raskar
In this course, we address the theoretical aspects of curl, divergence and integrability of vector fields, relevant to vision and graphics problems. We discuss scenarios where it is beneficial to operate on gradients than image intensities for image understanding, manipulation and synthesis. We review gradient domain techniques; address issues involved in 2D and 3D reconstructions from gradients, discuss implementations/numerical methods and give in-depth technical insight into the modern applications that exploit gradient domain manipulations.
Tensor Methods for Computer Vision, Graphics and Machine Learning
Lecturers: M. Alex O. Vasilescu, Amnon Shashua
Tensor factorizations of higher order tensors have been successfully applied in numerous machine learning, vision, graphics and signal processing tasks in recent years and are drawing a lot of attention. There are two main types of higher order tensor decompositions which generalize different concepts of the matrix SVD, the rank-R decomposition (open problem) and Rank-(R1,R2,...,RM) decomposition, plus various tensor factorizations under convex constraints relevant to classical inference and clustering tasks. The tutorial will provide an introduction to these techniques and show applicationd to compression, face recognition, multi-object detection in supervised and unsupervised settings, gait recognition, and computer graphics.
Oct. 15th, Morning
Content-based image and video retrieval
Lecturers: Theo Gevers, Nicu Sebe, Arnold Smeulders
In this tutorial, we give a survey of the most recent developments on image and video search engines. First, the important step of feature extraction will be discussed in detail such as color, shape and texture information, particularly paying attention to discriminatory power and invariance. Then, we focus on the concepts of indexing and genre classification as intermediate step to sort the data. We pay attention to (interactive) ways to perform browsing and retrieval by means of information visualization and relevance feedback.
Principles of Appearance Acquisition and Representation
Lecturers: Tim Weyrich, Jason Lawrence, Hendrik P.A. Lensch, Szymon Rusinkiewicz, Todd Zickler
Algorithms for scene understanding and realistic image synthesis require accurate models of the way real-world materials scatter light. The course describes recent work in both the vision and graphics communities to measure the spatially- and directionally-varying reflectance and subsurface scattering of complex materials, and to develop efficient representations and analysis tools for these datasets. It covers the design of acquisition devices and capture strategies for BRDFs and BSSRDFs, efficient factored representations, and a case study of capturing the appearance of human faces.
Oct. 15th, Afternoon
Lecturers: Jiri Matas, Krystian Mikolajczyk
The state state of the art in visual recognition will be presented via selected case studies. Next, key components and algorithms will be analysed and compared in depth: learning methods, matching and search algorithms, object representation, indexing approaches. Finally, "forgotten" and open problems will be reviewed, such as recognition of objects without surface texture, of wire-like and semi-transparent objects etc.
Discrete Optimization methods in Computer Vision
Lecturers: Nikos Komodakis, Philip Torr, Vladimir Kolmogorov, Yuri Boykov
Discrete Markov Random Fields (MRFs) can model a wide variety of problems in computer vision and pattern recognition. For this reason, MRF optimization is considered to be a task of fundamental importance, which has attracted a significant amount of research over the last years. The goal of this course will be to provide an overview for some of the recent developments in the field of MRF optimization. To this end, state-of-the-art discrete optimization algorithms for MRFs will be reviewed, while, in addition, the underlying ideas and principles behind these methods will be explained.