Loading…
Drag to turn the head (up to 15°) · scroll or pinch to zoom
GaussianTalker · WebGPU
Text-to-speech (a Piper voice fine-tuned on the person, or a personalised Kokoro voice), audio features (DeepSpeech, or the TTS's own text features), audio-driven Gaussian deformation and splat rendering, all running in your browser, on the GPU (WebGPU) or the CPU.
loading…
- GPU
- –
- Renderer
- –
- Gaussians
- –
- Render
- – · latency – · dropped 0
- Pose
- –
- Audio
- –
- TTS
- –
- Features
- –