Visual speech recognition using PCA networks and LSTMs in a tandem GMM-HMM system

Keyphrases

• Visual Speech Recognition 100%
• PCA Network 100%
• Gaussian Mixture Model-hidden Markov Model (GMM-HMM) 100%
• Principal Coordinate Analysis (PCoA) 50%
• Challenging Tasks 50%
• Recurrent Neural Network 50%
• Pattern Recognition 50%
• Audio Data 50%
• Video Data 50%
• Working Principle 50%
• Learned Features 50%
• Convolutional Networks 50%
• Speech Recognition 50%
• Unsupervised Learning 50%
• Noisy-OR 50%
• Audiovisual Archives 50%
• Image Patch 50%
• Spatio-temporal Features 50%
• Frontal View 50%
• Long-short Term Memory Cell 50%
• Utterance 50%
• Block Histogram 50%
• Phrase Recognition 50%

Computer Science

• Long Short-Term Memory Network 100%
• Speech Recognition 100%
• Recurrent Neural Network 33%
• Principle Component Analysis 33%
• Convolutional Network 33%
• Unsupervised Learning 33%
• Applied Technique 33%
• Pattern Recognition 33%