Audio2Face-3D

Esta página da Soniqo documenta Audio2Face-3D conforme a implementação local em speech-swift / speech-core. Os links do Hugging Face ficam abaixo das notas de integração.

Página interna primeiro

Cards e menus apontam primeiro para cá; os links do modelo fonte e dos bundles continuam disponíveis nesta página.

Visão geral

Modelo	Audio2Face-3D
Papel	Speech-to-avatar motion coefficients for downstream renderers
Backend	MLX
Saída	JSONL frames: timestamp plus the full coefficient vector
Idiomas	Language-independent, audio-driven
Licença	NVIDIA Open Model License
Status	Ready through speech avatar-motion and the Audio2Face3D Swift product
Fonte	NVIDIA Audio2Face-3D
Produto Swift	`Audio2Face3D`
CLI / runtime	`speech avatar-motion`

Uso

O trecho abaixo reflete a API ou comando atual exposto por 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-MLX

Links do modelo

Notas de implementação

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.