1. Frederico Rodrigues and Isabel Trancoso INESC/IST, 2000 Robust Recognition of Digits and Natural Numbers

2. 2 Summary Problem overview Baseline system Extensions to the baseline system Conclusions and future work

3. 3 The Problem Speaker Gender Age Vocal tract characteristics Pronunciation Rate of Speech Stress Lombard Reflex Microphone Position Distortion Channel Distortion Noise Environment Background noises Intermitent noises Coktail party noises Reverberation

4. 4 Corpus Description Multilingual telephone speech corpus SPEECHDAT(M)1000 speakers SPEECHDAT(II)4000 speakers Orthographically transcribed including noise events

5. 5 Noise events  [spk]:Speaker related noises  [sta]:Stationary noises  [int]:Intermittent noises

7. 7 Train and Test Set Definition Selection procedure –Age, gender and region distribution are approximately equal in both train and test sets; SPEECHDAT II –Fixed 500 speakers evaluation set –Additional 300 speakers development set SPEECHDAT(M) –200 speakers evaluation set Overall ratio of 80% Train/20% Test

8. 8 Sub-corpus Used I1 - Isolated digit strings B1 - Sequences of 10 digits N* - Natural numbers

9. 9 Feature Extraction MFCC (Mel Frequency Cepstral Coefficients) –14 Cepstra + 14  Cepstra + Energy +  Energy –Speech signal band-limited between 200 and 3800 Hz –Hamming Window: 25 ms each 10 ms Cepstral Mean Substraction –Simple but effective technique for channel and speaker normalization

10. 10 Acoustic Modeling Left-right continuous density HMM’s –Word models for each digit. No skips. –Silence and filler models with forward and backward skips Gender dependent models HMM: Hidden Markov Model

11. 11 Model Topology Fillers and silence models topology

12. 12 Baseline System - Isolated Digits Choose isolated digits with no noise marks –HMM parameters initialized with the global mean and variance of the training data Embedded Baum-Welch Reestimation Evaluate performance withViterbi decoding –Grammar allowing one digit and initial and final silence –Grammar allowing one digit and any number of fillers or silence

13. 13 Baseline System - Isolated Digits

14. 14 Baseline System - Isolated Digits Increment Gaussian mixtures per state up to 3 for the digit models Introduce files with noise marks Repeat re-estimation/evaluation process Increment Gaussian mixtures per state up to 3 for the filler and digit models

15. 15 Connected vs Isolated Digits Example: Number 3 1 2 6 said as: Isolated Digits: t r e S u~ d o j S s 6 j S Connected Digits: t r e z u~ d o j S _ 6 j S

16. 16 Baseline System - Connected Digits Use best isolated digit models as bootstrap models Repeat re-estimation/evaluation process Increment gradually Gaussian mixtures per state up to 5 for the digit models

17. 17 Baseline System - Results

18. 18 Extension to the Baseline System New way of modelling the filler models Same training/evaluation process Train the 9 filler and silence models with no skips Build a unique filler model concatenating all filler and silence models

19. 19 New Filler Model Arquitecture

20. 20 Results With New Filler Model

21. 21 Natural Numbers Phone models with 3 states and no skips Larger vocabulary size May be adapted to other tasks Phones initialized from models already trained for a directory assistance task Digits are still modeled by word models Grammar for natural numbers ranging from zero to hundreds of millions

22. 22 Natural Numbers Example Number 25: Hypothesis 1: vinte e cinco (Twenty and five) Hypotesis 2: vinte cinco (Twenty five) But “vinte cinco” could also be the sequence of natural numbers: 20 5

23. 23 Natural Numbers - Results

24. 24 Sample Application State Control Speech Recording User Server Feature Extraction Speech RecognitionSpeech Synthesis DIXI - SVIT Client Speech Prompts Speech / Commands Synthesised answer/ Commands Answer

25. 25 Conclusions and Future Work Explicitly modeling fillers is a difficult task –Improved filler model decreases error rate up to 50 % Develop context dependent models –Solve vowel reduction and co-articulation problems Results may be improved through the use of discriminative training techniques