Network Training for Continuous Speech Recognition

1
Network Training for Continuous Speech Recognition

• Author
• Issac John Alphonso
• Inst. for Signal and Info. Processing
• Dept. Electrical and Computer Eng.
• Mississippi State University
• Contact Information
• Box 0452
• Mississippi State University
• Mississippi State, Mississippi 39762
• Tel [phone]
• Fax [phone]

Email alphonso_at_isip.msstate.edu

• URL isip.msstate.edu/publications/books/msstate_t
  heses/2003/network_training/

2
INTRODUCTION
ORGANIZATION

• Motivation Why do we need a new training
  paradigm?
• Theoretical Review the EM-based supervised
  training framework.
• Network Training The differences between the
  network training and traditional training.
• Experiments Verification of the approach using
  industry standard databases (e.g., TIDigits,
  Alphadigits and Resource Management).

Motivation
Network Training
Experiments
Conclusions
3
INTRODUCTION
MOTIVATION

• A traditional trainer uses an EM-based framework
  to estimate the parameters of a speech
  recognition system.
• EM-based parameter estimation is performed in
  several complicated stages which are prone to
  human error.
• A network trainer reduces the complexity of the
  training process by employing a soft decision
  criterion.
• A network trainer achieves comparable performance
  and retains the robustness of the EM-based
  framework.

4
NETWORK TRAINER
TRAINING RECIPE
CI Training
State Tying
Flat Start
CD Training
Context-Independent
Context-Dependent

• The flat start stage segments the acoustic signal
  and seed the speech and non-speech models.
• The context-independent stage inserts and
  optional silence model between words.
• The state-tying stage clusters the model
  parameters via linguistic rules to compensate for
  sparse training data.
• The context-dependent stage is similar to the
  context-independent stage (words are modeled
  using context).

5
NETWORK TRAINER
FLEXIBLE TRANSCRIPTIONS
Network Trainer
SILENCE
HAVE
SILENCE

• The network trainer uses word level
  transcriptions which does not impose restrictions
  on the word pronunciation.
• The traditional trainer uses phone level
  transcriptions which uses the canonical
  pronunciation of the word.
• Using orthographic transcriptions removes the
  need for directly dealing with phonetic contexts
  during training.

6
NETWORK TRAINER
FLEXIBLE TRANSCRIPTIONS
Network Trainer
Traditional Trainer

• The network trainer uses a silence word which
  precludes the need for inserting it into the
  phonetic pronunciation.
• The traditional trainer deals with silence
  between words by explicitly specifying it in the
  phonetic pronunciation.

7
NETWORK TRAINER
DUAL SILENCE MODELLING
Multi-Path
Single-Path

• The multi-path silence model is used between
  words.
• The single-path silence model is used at
  utterance ends.

8
NETWORK TRAINER
DUAL SILENCE MODELLING

• The network trainer uses a fixed silence at
  utterance bounds and an optional silence between
  words.
• We use a fixed silence at utterance bounds to
  avoid an underestimated silence model.

9
NETWORK TRAINER
DUAL SILENCE MODELLING

• Network training uses a single-path silence at
  utterance bounds and a multi-path silence between
  words.
• We use a single-path silence at utterance bounds
  to avoid uncertainty in modeling silence.

10
EXPERIMENTS
TIDIGITS WER COMPARISON

• The network trainer achieves comparable
  performance to the traditional trainer.
• The network trainer converges in word error rate
  to the traditional trainer.

11
EXPERIMENTS
AD WER COMPARISON

• The network trainer achieves comparable
  performance to the traditional trainer.
• The network trainer converges in word error rate
  to the traditional trainer.

12
EXPERIMENTS
RM WER COMPARISON

• The network trainer achieves comparable
  performance to the traditional trainer.
• It is important to note that the 1.8 degradation
  in performance is not significant (MAPSSWE test).

13
CONCLUSIONS
SUMMARY

• Explored the effectiveness of a novel training
  recipe in the reestimation process of for speech
  processing.
• Analyzed performance on three databases.
• For TIDigits, at 7.6 WER, the performance of the
  network trainer was better by about 0.1.
• For OGI Alphadigits, at 35.3 WER, the
  performance of the network trainer was better by
  about 2.7.
• For Resource Management, at 27.5 WER, the
  performance degraded by about 1.8 (not
  significant).