Speaker-Independent Silent Speech Recognition from Flesh-Point Articulatory Movements Using an LSTM Neural Network

Mathematics

• Speech Recognition 100%
• Movement 66%
• Neural Networks 61%
• Memory Term 37%
• Recurrent Neural Networks 36%
• Normalization 27%
• Normalize 12%
• Feature Space 12%
• Speech 12%
• Training 9%
• Convert 9%
• Person 9%
• Experimental Results 9%
• Model 9%
• Linear regression 9%
• Text 8%
• History 8%
• Maximum Likelihood 8%
• Scaling 8%
• Minimise 7%
• Dependent 6%
• Range of data 5%
• Performance 5%
• Standards 5%

Engineering & Materials Science

• Speech recognition 61%
• Neural networks 39%
• Long short-term memory 32%
• Recurrent neural networks 29%
• Linear regression 9%
• Deep neural networks 9%
• Maximum likelihood 8%

Physics & Astronomy

• speech recognition 99%
• tongue 12%
• regression analysis 9%
• education 9%
• histories 7%
• electromagnetism 6%
• scaling 6%
• performance 4%