Vid2speech: Speech reconstruction from silent video

Keyphrases

• Convolutional Neural Network 100%
• Speech Restoration 100%
• Speech Signal 33%
• Acoustics 33%
• Learnability 33%
• Frame-based 33%
• Sound Features 33%
• Video Frames 33%
• Out-of-vocabulary Words 33%
• Speech Features 33%
• Speechreading 33%
• Word Intelligibility 33%
• Intelligible Speech 33%
• Automatic Feature Learning 33%

Computer Science

• Convolutional Neural Network 100%
• Representation Learning 33%
• Neighboring Frame 33%