Header menu link for other important links
X
Improving acoustic models in TORGO dysarthric speech database
Published in Institute of Electrical and Electronics Engineers Inc.
2018
PMID: 29522408
Volume: 26
   
Issue: 3
Pages: 637 - 645
Abstract
Assistive speech-based technologies can improve the quality of life for people affected with dysarthria, a motor speech disorder. In this paper, we explore multiple ways to improve Gaussian mixture model and deep neural network (DNN) based hidden Markov model (HMM) automatic speech recognition systems for TORGO dysarthric speech database. This work shows significant improvements over the previous attempts in building such systems in TORGO. We trained speaker-specific acoustic models by tuning various acoustic model parameters, using speaker normalized cepstral features and building complex DNN-HMM models with dropout and sequence-discrimination strategies. The DNN-HMM models for severe and severe-moderate dysarthric speakers were further improved by leveraging specific information from dysarthric speech to DNN models trained on audio files from both dysarthric and normal speech, using generalized distillation framework. To the best of our knowledge, this paper presents the best recognition accuracies for TORGO database till date. © 2001-2011 IEEE.
About the journal
JournalData powered by TypesetIEEE Transactions on Neural Systems and Rehabilitation Engineering
PublisherData powered by TypesetInstitute of Electrical and Electronics Engineers Inc.
ISSN15344320
Open AccessNo
Concepts (53)
  •  related image
    Database systems
  •  related image
    Deep neural networks
  •  related image
    Distillation
  •  related image
    Gaussian distribution
  •  related image
    Hidden markov models
  •  related image
    Multitasking
  •  related image
    Speech
  •  related image
    Trellis codes
  •  related image
    Acoustic model
  •  related image
    Automatic speech recognition system
  •  related image
    BUILDING COMPLEXES
  •  related image
    CEPSTRAL FEATURES
  •  related image
    DYSARTHRIA
  •  related image
    GAUSSIAN MIXTURE MODEL
  •  related image
    Recognition accuracy
  •  related image
    SPECIFIC INFORMATION
  •  related image
    Speech recognition
  •  related image
    Acoustics
  •  related image
    Article
  •  related image
    Artificial neural network
  •  related image
    DYSARTHRIA
  •  related image
    Entropy
  •  related image
    Hidden markov model
  •  related image
    Human
  •  related image
    MOTOR PERFORMANCE
  •  related image
    SPEECH ARTICULATION
  •  related image
    Speech intelligibility
  •  related image
    Adult
  •  related image
    Automatic speech recognition
  •  related image
    COMMUNICATION AID
  •  related image
    Complication
  •  related image
    DYSARTHRIA
  •  related image
    Equipment design
  •  related image
    Factual database
  •  related image
    Female
  •  related image
    Machine learning
  •  related image
    Male
  •  related image
    Markov chain
  •  related image
    Normal distribution
  •  related image
    SELF HELP DEVICE
  •  related image
    Speech analysis
  •  related image
    SPEECH DISORDER
  •  related image
    Statistical model
  •  related image
    COMMUNICATION AIDS FOR DISABLED
  •  related image
    Databases, factual
  •  related image
    DYSARTHRIA
  •  related image
    Humans
  •  related image
    Markov chains
  •  related image
    Models, statistical
  •  related image
    SELF-HELP DEVICES
  •  related image
    SPEECH DISORDERS
  •  related image
    SPEECH PRODUCTION MEASUREMENT
  •  related image
    SPEECH RECOGNITION SOFTWARE