AUTOMATIC SPEECH RECOGNITION IN PRESENCE OF MUSIC NOISE ON MULTICHANNEL FAR-FIELD RECORDINGS
Annotation
Subject of Research. The paper considers a method of music noise reduction in a multichannel speech signal based on noise mask estimation. The method is applied for automatic speech recognition in presence of music noise. Method. The study is performed using an acoustic model implemented in artificial neural networks and real life recordings performed in reverberant conditions. Main Results. It is shown that the acoustic model is capable of estimating the noise mask on a multichannel mixture for different music genres. The application of such mask to covariance matrix estimation for MVDR (Minimum Variance Distortionless Response) beamforming algorithm results in increasing the recognition accuracy by at least 4.9 % at signal-noise ratio levels of 10–30 dB. Practical Relevance. The method of MVDR coefficient estimation based on noise mask estimation by an acoustic model serves to suppress non-stationary noise, such as music, thus increasing the robustness of automatic speech recognition systems.
Keywords
Постоянный URL
Articles in current issue
- LOW-COHERENCE REFLECTOMETRY OF FLUORESCENT RANDOM MEDIA
- MODELING OF INTEGRATED OPTICAL QUANTUM SEARCH ALGORITHM
- APPLICATION OF INFRARED SPECTROSCOPY AND MULTIVARIANT ANALYSIS TO STUDY OF SERUM FOR PATIENTS WITH EPILEPSY
- SEARCH METHOD FOR CHANGES OF THE EARTH’S SURFACE STATE THROUGH MULTI-TEMPORAL SATELLITE IMAGES
SILICON SURFACE MICROSTRUCTURING BY SINGLE-EXPOSURE FEMTOSECOND DOUBLE LASER PULSE
MODIFIED BACKSTEPPING ALGORITHM FOR CONTROL OF NONLINEAR SYSTEMS WITH CROSS-COUPLINGS
COMPARISON OF UNKNOWN PARAMETERS ESTIMATES BY METHOD OF DYNAMIC REGRESSOR EXTENSION AND MIXING AND LEAST SQUARE METHOD IN NOISE PRESENCE
- AMMONIUM SULPHATE EFFECT ON CHARACTERISTICS OF YAG:Yb NANOPOWDERS AND OPTICAL CERAMICS
- ENHANCEMENT OF Eu3+ PHOTOLUMINESCENCE IN SODIUM-ALUMINOSILICATE GLASSES BY SILVER MOLECULAR CLUSTERS FORMED WITH Na+–Ag+ ION EXCHANGE METHOD
- MODELING OF ZnO ELECTRONIC STRUCTURE FROM FIRST PRINCIPLES BY APPLYING ADVANCED FUNCTIONALS
- ADAPTIVE MODULE DEVELOPMENT FOR CREATION AND STUDY OF VIRTUAL MODELS OF ENVIRONMENTAL OBJECTS
NONSTATIONARY PROCESSES PERIOD ESTIMATION IN CLOUD SYSTEMS
- ADAPTIVE THREE-DIMENSIONAL DISCRETE COSINE TRANSFORM OF TRANSPORT IMAGES
- PROCESSING OF SIGNAL INFORMATION IN PROBLEMS OF MONITORING INFORMATION SECURITY OF UNMANNED AUTONOMOUS OBJECTS
- INTELLIGENT TOURIST ASSISTANCE SYSTEM: SERVICE-ORIENTED ARCHITECTURE AND IMPLEMENTATION
- AUTOMATIC HYPERPARAMETER OPTIMIZATION FOR CLUSTERING ALGORITHMS WITH REINFORCEMENT LEARNIN
- PARAMETRIC OPTIMIZATION OF DIGITAL INTEGRATED CIRCUITS FOR MICROMECHANICAL SENSORS
PROCESS-ORIENTED SYNTHESIS OF SUCCESSIVE APPROXIMATION ANALOG-TO-DIGITAL CONVERTERS FOR INTEGRATED CIRCUITS
- 3D-MODELING OF QUARTZ GLASS SENSORY ELEMENTS OF HEMISPHERICAL RESONATOR GYRO AND PENDULUM ACCELEROMETER
- THERMAL MODE OF ULTRACOLD NEUTRON SOURCE AT WWR-M REACTOR
- EFFECT OF VARIOUS DIMENSION CONVOLUTIONAL LAYER FILTERS ON TRAFFIC SIGN CLASSIFICATION ACCURACY
- INFORMATION SUPPORT OF DECISION-MAKING IN COMPUTER-AIDED RELIABILITY ORIENTED DESIGN
- AUTOMATIC SPEECH RECOGNITION IN PRESENCE OF MUSIC NOISE ON MULTICHANNEL FAR-FIELD RECORDINGS
- REVIEW ON SET OF SCIENTIFIC WORKS BY TERTYCHNY-DAURI V.YU. GALAMEKH. INSIXVOLUMES. THE 2nd REVISEDEDITION