Speech inpainting: Context-based speech synthesis guided by video

Computer Science

• Visual Modality 100%
• Speech Synthesis 100%
• Visual Content 50%
• Visual Feature 50%
• Deep Learning Model 50%
• Speech Acoustics 50%
• Speech Recognition 50%
• Facial Expression 50%
• Transformer-Based Deep Learning 50%

Keyphrases

• Audio Modality 33%
• AV-HuBERT 33%
• Transformer-based Deep Learning 33%
• Audio-visual Models 33%