Acoustical and environmental robustness in automatic speech recognition
Acoustical and environmental robustness in automatic speech recognition
Multiple approaches to robust speech recognition
HLT '91 Proceedings of the workshop on Speech and Natural Language
Reduced channel dependence for speech recognition
HLT '91 Proceedings of the workshop on Speech and Natural Language
Modified MFCC windowed technique for speaker word recognition
Proceedings of the International Conference & Workshop on Emerging Trends in Technology
| Hi-index | 0.00 |
With the intention of developing a robust speech recognizer largely immune to the vagaries of extrinsic changes, we investigated the consequences of various background noises and microphones on the performance of our Tangora system. We identified several noisy locations such as our cafeteria and our secretary's office and included a relatively quiet office for comparison. We recorded isolated-word training and test data from one male and one female speaker at different locations employing several varieties of microphones. A typical experiment consisted of designing a speaker-dependent HMM system with one set of training data and decoding the test data collected at all locations. We found that microphone characteristics had a significant impact on the robustness of our system. Another observation was that controlled contamination of the quiet training data with ambient noise improved the noise immunity of the recognizer, discounting the role of Lombard effect in our studies.