Increase of probability of signal recognition rate according to their cross-correlation portraits
- 1 д.т.н., профессор, зав. кафедрой прикладной математики и информатики, Ульяновский государственный технический университет
- 2 к.т.н., доцент, зав. кафедрой английского языка для профессиональной деятельности, Ульяновский государственный университет
- 3 аспирантка, Ульяновский государственный технический университет
Abstract
The problem of signal recognition under intensive noise conditions appear in many practically important applications. For example, such a problem can be found in process monitoring and speech information-control systems. All these systems must function well under intensive noise conditions. If there is a fixed number (library) of recognized signals it is necessary to conduct recognition by comparing signals with their models. In order to improve recognition the signals are previously transformed into special images (cross-correlation portraits). However, inexact determination of speech signal boundaries and their variability have a negative impact. The paper suggests several methods to reduce impact of inexact determination by estimation variability of speech signal boundaries. The optimization of model library helps to reflect signal variability better. As a result the probability of recognition rate increases.
How to cite
Крашенинников В.Р., Крашенинникова Н.А., Галицкая Е.Ю. Повышение вероятности правильного распознавания сигналов по их кросскорреляционным портретам // Радиотехника. 2014. № 7. С. 107–110.
References
- Krasheninnikov V.R., Armer A.I., Kuznetsov V.V., Lebedeva E.Yu. Cross-Correlation Portraits of Voice Signals in the Problem of Recognizing Voice Commands According to Patterns // Pattern Recognition and Image Analysis. 2011. V. 21. № 2. P. 185-187.
- Klyachkin V.N. Modeli i metody' statisticheskogo kontrolya mnogoparametricheskogo texnologicheskogo proczessa. M.: Fizmatlit. 2011. 196 s.
- Vasil'ev K. K., Krasheninnikov V.R. Statisticheskij analiz mnogomerny'x izobrazhenij. Ul'yanovsk: UlGTU. 2007. 170 s.
- Vasil'ev K. K., Krasheninnikov V.R. Adaptivny'e algoritmy' obnaruzheniya anomalij na posledovatel'nosti mnogomerny'x izobrazhenij // Komp'yuternaya optika. 1995. Vy'p. 14-15. Ch. 1. S. 125-132.
- Krasheninnikov V.R., Potapov M.A. Estimation of Parameters of Geometric Transformation of Images by Fixed-Point Method // Pattern Recognition and Image Analysis. 2012. V. 22. № 2. P. 303-317.
- Vasil'ev K. K., Krasheninnikov V.R., Tashlinskij A.G. Statisticheskij analiz posledovatel'nostej mnogomerny'x izobrazhenij // Naukoyomkie texnologii. 2013. T. 14. № 5. S. 5-11.
- Tashlinskij A.G., Kaveev I.I., Voronov S.V. Metodika privyazki izobrazhenij v usloviyax intensivny'x pomex // Radiotexnika. 2012. № 9. S. 45-49.
- Krasheninnikov V.R., Armer A.I., Krasheninnikova N.A., Xvostov A.V. Raspoznavanie rechevy'x komand na fone intensivny'x pomex s pomoshh'yu avtokorrelyaczionny'x portretov // Naukoemkie texnologii. Moskva. 2007. № 9. S. 65-76.
- Krasheninnikov V.R., Lebedeva E.Ju., Kapy'rin V.K. Variacziya granicz rechevy'x komand dlya uluchsheniya raspoznavaniya rechevy'x komand po ix krosskorrelyaczionny'm portretam // Izv. Samarskogo nauchnogo czentra Rossijskoj akademii nauk. 2013. T. 15. № 4-4. S. 928-930.
- Krasheninnikova N.A. Forming a library of patterns for recognizing verbal commands // Pattern Recognition and Image Analysis. 2009. V. 19. № 1. P. 48-51.
- Krasheninnikov V.R., Krasheninnikova N.A., Kuznetsov V.V., Lebedeva E.Yu. Optimization of dictionary and model library for recognition of speech commands // Pattern Recognition and Image Analysis. 2011. V. 21. № 3. P. 505-507.
- Krasheninnikov V.R., Krasheninnikova N.A., Kuznetsov V.V., Lebedeva E.Yu. Optimization of Dictionary and Model Library for Recognition of Speech Commands Based on Cross-Correlation Portraits // Pattern Recognition and Image Analysis. 2013. V. 23. № 1. P. 80-86.