| ECVision indexed and annotated bibliography of cognitive computer vision publications This bibliography was created by Hilary Buxton and Benoit Gaillard, University of Sussex, as part of ECVision Specific Action 8-1 The complete text version of this BibTeX file is available here: ECVision_bibliography.bib
G. Medioni and M.-S. Lee and C.-K. Tang
ABSTRACT
The design and implementation of a complete artificial vision system represents a major challenge. There is still a wide gap between the state of the art and the goal. Researchers have not been able to cleanly decompose and express the sub - problems to be addressed. Therefore, there are many drawbacks for the classification of approaches into low, medium, and high level vision: o Low-level modules, such as edge detectors, produce primitives that {"}should be corrected and improved by higher level modules{"}. o High-level modules, such as 3 -D shape inference, or behavior analysis, work remarkably well on perfect data, but degrade abruptly with real data. o Mid-level modules are supposed to bridge the gap between low and high levels, and as such, get a long list of tasks, and a good share of the blame for failure. The book represents a summary of the research that has been conducted since the early 1990s. It describes a conceptual framework, which addresses some current shortcomings, and proposes a unified approach for a broad class of problems. The main theme of the book is to present the elements of the author’s “salient feature interference engin e, and theirinteraction.” The book also introduces tensors as a way to represent information, tensor fields as a way to encode both constraints and results, and tensor voting as the communication scheme. |
|
| Site generated on Friday, 06 January 2006
|