Language Identification

1
Language Identification

• Oldrich Plchot, Pavel Matejka Speech_at_FIT, Brno
  University of Technology, Czech Republic
• matejkap_at_fit.vutbr.cz

IKR Brno 2012
2
Outline

• Why do we need LID?
• Evaluations
• Acoustic LID
• Phonotactic LID
• Fusion
• Conclusion

3
Why do we need language identification?

• 1) Route phone calls to human operators.

Emergency (112,155,911)
Call centers
Fireguard (150)
Police (158)
4
Why do we need language identification?

• 2) Pre-select suitable recognition system.

KWS CHN
Language Identification
Translate SPA
Translate CZE
Translate VIE
Speech2Text ENG
Connect
5
Why do we need language identification?

• 3) Security applications to narrow search space.

6
Two main approaches to LID

• Acoustic Gaussian Mixture Model

• Phonotactic Phoneme Recognition followed by
  Language Model

7
Acoustic approach

• Gaussian Mixture Model

• good for short speech segments and dialect
  recognition
• - relies on the sounds

8
Spectral features - MFCC
20ms
10ms
-12.8 -0.3 -5.7 -22.4 8.9 6.8
-11.2 0.4 -4.7 -13.0 2.3 4.5
Short-time FFT
Mel - Filter Bank
Log ()
Discrete Cosine Transform
9
Shifted delta cepstra

• Shifted Delta Cepstra represent an information
  about the speech evolution around the current
  frame ( 0.1sec)
• Size of Final feature vector is 7 MFCC 7 7
  SDC 56

10
Acoustic systems GMM based

• Maximum likelihood (generative)
• Objective function to maximize is the likelihood
  of training data given the transcription
• Maximum Mutual Information (discriminative)
• Objective function to maximize is the posterior
  probability of all training utterances being
  correctly recognized
• Advantages of using discriminative training
• Lower error rates
• Less parameters
• Disadvantages of discriminative training
• Overtraining
• Sometimes computationaly expensive
• Channel Compensation from previous presentation

11
Highly overlapped distributions
12
Results on LRE 2007 (14 languages)
System / Equal Error Rate 30sec 10sec 3sec
GMM2048 ML 8.03 12.89 21.77
GMM 256 ML 16
GMM256 MMI (15 MMI iterations) 4.15 8.61 18.43
GMM256-MMI-chcf (3 MMI iterations) 3.73 9.81 20.98
System / Equal Error Rate 30sec 10sec 3sec
GMM2048 8.03 12.89 21.77
GMM2048-eigchan 2.76 7.38 17.14
GMM2048chcf 2.94 7.40 17.93
GMM2048-MMI-chcf ( 3 MMI iterations) 2.41 7.02 16.90

• The best acoustic system combines
• Many Gaussians
• Eigen-channel compensation of features
• MMI

13
Phonotactic approach

• Phoneme Recognition followed by Language Model
  (PRLM)

• good for longer speech segments
• robust against dialects in one language
• eliminates speech characteristics of speaker's
  native language

14
Phone recognizer

• Investigation of different phone recognizers for
  LID gt better phone recognizer better LID system

• 3 neural networks to produce the phone posterior
  probability
• 310 ms long time trajectory around the actual
  frame

15
Phone recognition output
One best phone string
16
Phonotactic modeling - example
German
English
Test
u n d 25
a n d 3
t h e 0
. . . .
u n d 1
a n d 32
t h e 13
. . . .
u n d 5
a n d 0
t h e 1
. . . .

• N-gram language models discounting, backoff
• Support Vector Machines vectors with counts
• PCA LDA
• Neural Networks

17
Phone recognition output
One best phone string
Phone lattice
18
Results on LRE 2007 (14 languages)
System / Equal Error Rate 30 sec 10 sec 3 sec
HU_LM string (4-gram) 6.35 13.86 27.12
HU_LM 5.54 11.75 23.54
HU_SVM-3gram-counts 5.41 13.26 26.92

• Conclusion
• Build as good phone recognizer as you can
• Gather as much data for each language as you can
• Different approaches to modeling counts seem to
  not have big influence on results

19
Fusion - LRE 2007 (14 languages)
System / Equal Error Rate 30 sec 10 sec 3 sec
Acoustic - GMM2048-MMI-chcf ( 3 MMI iterations) 2.41 7.02 16.90
Phonotactic - EN_TREE 3.54 10.68 22.66
Phonotactic - HU_TREE_A3E7M5S3G3_LFA 4.52 10.35 23.66
Fusion The best 3 systems 1.28 4.63 13.53

• Note
• Fusion weights have to be trained on separate set
  of files which are as close as possible to target
  data

20
Thanks for your attention and I hope you enjoyed
it )