Projects / Programmes
METROLOGY AND BIOMETRIC SYSTEMS
January 1, 2009
- December 31, 2012
Code |
Science |
Field |
Subfield |
2.06.01 |
Engineering sciences and technologies |
Systems and cybernetics |
Control systems technology |
2.15.00 |
Engineering sciences and technologies |
Metrology |
|
Code |
Science |
Field |
P180 |
Natural sciences and mathematics |
Metrology, physical instrumentation |
Code |
Science |
Field |
2.02 |
Engineering and Technology |
Electrical engineering, Electronic engineering, Information engineering |
magnetics, biomagnetics, standard (etalon), digital signal processing, acoustics, speech recognition, speech synthesis, biometric personal authentication
Researchers (18)
no. |
Code |
Name and surname |
Research area |
Role |
Period |
No. of publicationsNo. of publications |
1. |
22480 |
PhD Samo Beguš |
Engineering sciences and technologies |
Researcher |
2009 - 2012 |
167 |
2. |
11805 |
PhD Simon Dobrišek |
Engineering sciences and technologies |
Researcher |
2009 - 2012 |
270 |
3. |
05692 |
PhD Dušan Fefer |
Engineering sciences and technologies |
Researcher |
2009 - 2012 |
588 |
4. |
29556 |
PhD Rok Gajšek |
Engineering sciences and technologies |
Junior researcher |
2009 - 2012 |
29 |
5. |
13144 |
PhD Stanislav Gruden |
Engineering sciences and technologies |
Researcher |
2009 - 2012 |
36 |
6. |
22479 |
PhD Melita Hajdinjak |
Natural sciences and mathematics |
Researcher |
2009 - 2012 |
126 |
7. |
03075 |
PhD Anton Jeglič |
Engineering sciences and technologies |
Researcher |
2009 - 2012 |
508 |
8. |
31978 |
PhD Tadej Justin |
Engineering sciences and technologies |
Junior researcher |
2009 - 2012 |
22 |
9. |
31985 |
PhD Janez Križaj |
Engineering sciences and technologies |
Junior researcher |
2009 - 2012 |
36 |
10. |
21321 |
PhD Aleš Mihelič |
Engineering sciences and technologies |
Researcher |
2009 - 2012 |
50 |
11. |
09580 |
PhD France Mihelič |
Engineering sciences and technologies |
Researcher |
2009 - 2012 |
313 |
12. |
01938 |
PhD Nikola Pavešič |
Engineering sciences and technologies |
Head |
2009 - 2012 |
656 |
13. |
34297 |
Martin Stanonik |
Engineering sciences and technologies |
Technical associate |
2011 - 2012 |
7 |
14. |
28458 |
PhD Vitomir Štruc |
Engineering sciences and technologies |
Researcher |
2009 - 2012 |
332 |
15. |
17127 |
Drago Tacar |
|
Technical associate |
2009 - 2012 |
57 |
16. |
11380 |
PhD Mario Žganec |
Engineering sciences and technologies |
Researcher |
2009 - 2012 |
98 |
17. |
12000 |
PhD Jerneja Žganec Gros |
Engineering sciences and technologies |
Researcher |
2009 - 2012 |
290 |
18. |
19236 |
PhD Janez Žibert |
Engineering sciences and technologies |
Researcher |
2009 - 2012 |
263 |
Organisations (3)
Abstract
Within the scope of the research programme we will develop new measurement procedures to widen the acredited range and lower the measurement uncertainty of the measurements, thus the laboratory will meet requirements needed by the the industry and research institutes.
We will test the separation of instantaneous and convolutive signal mixtures with the methods for blind source signal separation. The signals will be simulated or recorded in real environment. We will analyse the possibilities for accreditation in the field of acoustic measurements for which we have the measurement equipment and anechoic chamber.
In the frame of the research programme we will put our special attention to the further development of speech technologies for the Slovenian language, which will enable multimodal communications with modern inteligent systems in the Slovenian language and the feasibility to develop of personal authentication systems based on a combination of several human physical and/or behavioural characteristics.
Significance for science
MAGNETIC MEASUREMENTS The described research program is proving theoretical and research support for the measurements of DC magnetic flux density and magnetic field at low frequencies. Research results will enhance current knowledge in the field of magnetic measurements. SPEECH TECHNOLOGIES Our research in the field of speech technologies resulted in new insights and novel methodologies for performance evaluation of automatic speech recognition systems. Here we managed to incorporate the temporal dimension of events in the evaluation procedures itself. In the area of feature selection/extraction for the task of speaker-state recognition, we have shown how to integrate phone recognition techniques into the feature selection/extraction procedure and, thus, reduce the performance difference between the results of text-dependent and text-independent recognition scenarios. For knowledge presentation in automated dialogue systems we were able to extend the K-relation model to the D-relation model, called relational algebra with similarities, which supports similarity-based querying. Because our model (in contrast to other models) satisfies all the classical relational algebra identities, it paves the way for a new generation of database models and information systems. BIOMETRIC SECURITY SYSTEMS The research conducted in the field of biometric security systems led to novel non-parametric score normalization techniques that can be used in any biometric verification system regardless of the underlying modality. The proposed family of techniques builds upon the established parametric score normalization techniques, but ensures greater robustness and performance of the biometric systems using them. In the field of statistical template modeling and model comparison we proposed a simplified version of the probabilistic linear discriminant analysis, which was found to provide state-of-the-art results in terms of robustness as well as recognition performance.
Significance for the country
During the research programme, a lot of attention has been dedicated to the transfer of the applied results of the research into new or improved products for the Slovenian industry. In scope of the research programme, several areas of language technologies for the Slovenian language have been addressed and significant results have been achieved that can be used to improve the quality of language technology products. For example, based on the research programme results, improvements to the Slovenian text-to-speech synthesis engine Proteus are already in progress. In a similar way, the significant results achieved in the field of biometric technologies have been taken up by the programme partners in form of industrial research and experimental delevopment of multimodal biometric products, based on speech and face biometrics.
Most important scientific results
Annual report
2009,
2010,
2011,
final report,
complete report on dLib.si
Most important socioeconomically and culturally relevant results
Annual report
2009,
2010,
2011,
final report,
complete report on dLib.si