Thèse de doctorat : Université de Fribourg, 2008 ; no 1610.
|
Thèse de doctorat : Université de Fribourg, 2008 ; no 1587.
This thesis proposes a method for indexing and browsing archives of multimedia documents, and in particular meeting recordings, using printable documents and links. Existing systems for indexing and browsing multimedia data have four main limits. First, the indexing requires high-level abstractions extracted from multimedia documents, which is still an unsolved problem for rich media such as...
|
Thèse de doctorat : Université de Fribourg, 2006 ; no 1513.
This thesis addresses the recognition of physical and logical structures of complex documents, rich in variability. More precisely, we studied the evolution of models within an interactive context where the system gradually integrates the knowledge induced by the corrections of the user. We studied the features of the Arabic language and we designed a recognition system for this language. In a...
|
Thèse de doctorat : Université de Fribourg, 2005 ; Nr. 1488.
Document recognition is a research domain that doesn’t lose its relevance even in a world where documents are increasingly often available in an electronic form. Whereas some years ago, the goal of document recognition was to convert documents from paper into an electronic form, the problem is shifted more and more from pure recognition towards document understanding. This requires much more...
|
Thèse de doctorat : Université de Fribourg, 2006 ; no. 1534.
The phonographic record was the only way to store sounds until the introduction of magnetic tape in the early 1950s. Therefore there are huge collections of phonographic records, for example in radio stations and national sound archives. Such archives include pressed discs, which were produced in mass by record companies for commercial distribution, as well as direct cut discs obtained by the...
|
Thèse de doctorat : Université de Fribourg, 2006.
This thesis proposes a multimodal alignment framework that bridges the gap between static documents and spoken language. This alignment aims mainly at linking static documents with temporal data, in order to exploit the multi-level structure of documents for indexing multimedia recordings of events. This novel multimodal alignment method, largely described in this thesis, is applied on two...
|
Thèse de doctorat : Université de Fribourg, 2006 ; no. 1529.
This thesis investigates methods for building an efficient application system for the document-based automatic indexing and retrieval (DocMIR) of multimedia data captured from multimodal environments such as meetings, conferences, etc. Both empirical image processing, video segmentation methods and document analysis approaches are studied to bridge the gap between temporal data and static...
|
Thèse de doctorat : Université de Fribourg : 2000 ; no 1312.
This thesis addresses the question of content-based image retrieval (CBIR) in heterogeneous databases. In an analysis of the existing CBIR tools that was done at the beginning of this work, we have shown that there was room for improvement in three key areas: query form, image and query representation, and computation of similarity. This analysis led us to studying the usability of a method for...
|