We build technologies in support of bringing virtual human to life. Our core A.I. technologies strive for realism, naturalness, scalability, and performance. Under this rationale, our tech reduces by a large margin the cost and turnaround time in production caused by the need for building specialized models for each new avatar. This unlocks unlimited possibilities in rapid creation in our virtual human ecosystem.
Click to shift
Providing one image of a person, our Face Shifting technology can apply the facial characteristics to any face images or videos.
Robust to extreme head pose, lighting, and occlusion
High retention of facial expression
Natural and high fidelity
Driving any face with a voice recording, our Lip Synchronization technology synchronises lip, mouth, and jaw movement with the speech.
Supports 50+ languages
Robust to background noise
Works for both real and synthetic voices
From Sam’s voice
To AI-generated woman voice
Given a short voice recording, our Voice Cloning technology can swap the vocal attributes onto another speech recording.
Clones a voice with a recording as short as one minute
Cross languages, dialects, accents, and genders
Preserves original speech pattern