FaceDiffuser: Speech-Driven 3D Facial Animation Synthesis Using Diffusion

Keyphrases

• Facial Expression Synthesis 100%
• Speech-driven 100%
• 3D Facial Animation 100%
• Non-deterministic 66%
• Blendshapes 66%
• Diffusion Method 66%
• Publicly Available 33%
• Challenging Tasks 33%
• Nonverbal 33%
• State-of-the-art Techniques 33%
• Vertex-based 33%
• Facial Cues 33%
• Deep Learning Methods 33%
• Representation Model 33%
• Deep Learning Model 33%
• Speech Input 33%
• Facial Animation 33%
• HuBERT 33%
• Audio Input 33%
• Subjective Analysis 33%
• Speech-driven Facial Animation 33%
• Speech Representation 33%

Computer Science

• Facial Animation 100%
• Representation Model 20%
• Deep Learning Model 20%
• Deep Learning Method 20%