Quaero is a collaborative research and development program, centered at developing multimedia and multilingual indexing and management tools for professional and general public applications such as the automatic analysis, classification, extraction and exploitation of information.
The research aims to facilitate the extraction of information in unlimited quantities of multimedia and multilingual documents, including written texts, speech and music audio files, and images and videos.
Quaero was created to respond to new needs for the general public and professional use, and new challenges in multimedia content analysis resulting from the explosion of various information types and sources in digital form, available to everyone via personal computers, television and handheld terminals.
The consortium is composed of French and German public and private research organisations.
The Quaero consortium is coordinated by Thomson. The other large industrial organizations participating are Jouve and Exalead. Dedicated technology suppliers Bertin, LTU, Synapse and Vecsys will contribute and further develop top notch technologies in their respective business domains.
French and German public research institutes, coordinated by CNRS are CNRS (INIST, LIMSI, IMMI), INRIA, Institut Telecom, IRCAM, IRIT, LIPN, MIG-INRA, Joseph Fourier University, University of Karlsruhe and RWTH Aachen University.
Finally, the participation of public institutions BnF, DGA, Ina and LNE demonstrates the strong support of the public sector to the success of the program.
Quaero and development program is set to receive 99 million euros in aid from the French agency OSEO that will handle it. Consortium members will contribute an equivalent amount to reach an overall budget of approximately 200 million euros for innovative research projects.
A number of differences exist between Quaero and other R&D programmes in similar areas:
- First, the size and scope of the planned research work is an important characteristic of the programme. Quaero gathers several applications and base technologies with multiple links between them. Each application will use several technologies and each technology does potentially address several applications. To be capable to effectively mutualise resources and know-how and cross-fertilize between applications, this architecture requires a certain level of critical mass to be reached. The structure is capable to expand to other application projects that could benefit from the base technologies, or to other technology developments addressing new needs from the application projects.
- Second, the central role taken by the evaluation indicates the demand for scientific excellence. By comparing the Quaero developments to results produced elsewhere the evaluation infrastructure will make sure that the Quaero developments are at the state of the art or beyond. Compared to other research programmes the Quaero evaluation effort is very high. It is similar to the American campaigns conducted by DARPA.
- Third, resources allocated to the development of the large annoted Corpus is also highly differentiating. Increasing the size of the corpus is a prerequisite to significant technological improvement.
- Fourth, Quaero is neither only a research, nor only an industry programme, but a combination of both aspects. Technological innovation will feed the ambitions of the applications. Feed back of demonstrations and proofs of concepts by industrial partners will allow refining technological research directions. Promising application innovations are expected not only for industrial areas, but also for cultural and scientific areas.
The coexistence of the 4 elements is Quaero’s specificity. These elements are expected to allow Quaero to reach its goals.
People involved in SAMOVA team
SAMOVA team is involved in the central role taken by the evaluation.
Each person is responsible of the evaluation of one task at least.
- Régine André-Obrecht (task 6.2: Music content extraction; task 6.3: Music classification; task 6.4: Music similarity; Music structuring and summary)
- Hervé Bredin (task 11.4: Audio fingerprinting; task 11.6: Video fingerprinting)
- Jérôme Farinas (task 5.3: language recognition)
- Isabelle Ferrané (task 5.4: Dialog and vocal interface; task 10.2Multimodal content recognition)
- Philippe Joly (task 9.4: Event recognition; task 9.5: Face person detection and recognition in Video)
- Julien Pinquier (task 6.1: Audio segmentation, Analysis and Classification)
- Christine Sénac (task 10.3: Cross Modal person recognition)
- Start time: 01 January 2008
- End time: 31 December 2012