Conference program in pdf file, the Conference proceedings is published as Spinger LNCS 4679 on line [see Cover pages]
Many problems in computer vision and pattern recognition can be formulated in terms of probabilistic inference or optimization of energy functions. EMMCVPR 2007 will address the critical issues of representation, learning, and inference. The first major theme is how to represent visual problems in terms of probabilistic inference using sophisticated probability distributions defined over structured relational systems, such as graphs and generative grammars. The second theme is the development of efficient learning and inference algorithms using advanced techniques from statistics, computer science, and applied mathematics. A third theme is the use of datasets with groundtruth to act as benchmarks for evaluating algorithms and as a way to train learning algorithms. A fourth theme is the relation of high-level vision to more general cognitive processes, including functional and semantic level descriptions in real world scenes. Additional topics include issues such as image parsing, cue integration, and the trade-offs between bottom-up and top-down processing.
As with the previous editions (1997, 1999, 2001 , 2003 and 2005 ), the proceedings will be published by Springer in the Lecture Notes on Computer Science (LNCS) series.
The scientific program of EMMCVPR 2007 will include the presentation of invited plenary talks and contributed research papers. The workshop, which is sponsored by the International Association for Pattern Recognition (IAPR), will be held in EZhou,Hubei,China .
A list of relevant topics includes (but is not restricted to):
|
APPROACHES
|
APPLICATIONS
|
