Audio2Face-3D
Diese Soniqo-Seite dokumentiert Audio2Face-3D aus der lokalen speech-swift- / speech-core-Implementierung. Hugging-Face-Bundles sind nach den Integrationshinweisen verlinkt.
Zuerst interne Seite
Landing-Karten und Docs-Menüs führen zuerst hierher; Quellen- und Bundle-Links bleiben auf dieser Seite verfügbar.
Überblick
| Modell | Audio2Face-3D |
|---|---|
| Rolle | Speech-to-avatar motion coefficients for downstream renderers |
| Backend | MLX |
| Ausgabe | JSONL frames: timestamp plus the full coefficient vector |
| Sprachen | Language-independent, audio-driven |
| Lizenz | NVIDIA Open Model License |
| Status | Ready through speech avatar-motion and the Audio2Face3D Swift product |
| Quelle | NVIDIA Audio2Face-3D |
| Swift-Produkt | Audio2Face3D |
| CLI / Laufzeit | speech avatar-motion |
Verwendung
Das folgende Snippet entspricht der aktuellen API oder dem aktuellen Befehl aus speech-swift.
# Generate avatar motion coefficients from speech audio.
.build/release/speech avatar-motion input.wav --output motion.jsonl --verbose
.build/release/speech avatar-motion input.wav \
--model aufklarer/Audio2Face-3D-v2.3.1-Claire-MLXModelllinks
Implementierungsnotizen
- Three identities are published: James (default) and Claire at v2.3.1 with 169 coefficients, Mark at v2.3 with 301; skin coefficients live in the identity's own geometry basis, so renderers need the matching rig or a retarget projection.
- Two-branch regressor: a trimmed Wav2Vec2 audio encoder blended across transformer layers, plus a windowed-autocorrelation frequency branch, fused through emotion-conditioned strided convolutions.
- Each 8320-sample window of 16 kHz mono audio (hop 4160) emits one frame, conditioned on a 26-value emotion vector — 16 implicit plus 10 explicit values.
- The MLX graph runs the full NVIDIA model forward pass and is parity-checked against an ONNX fixture generated from the official release.