Demo Inference

Run offline lip-reading with metrics, latency, and device info.

Settings

ModelExpected text (optional)DecoderBeam width

LLM post-processing (Gemini)

Upload a video for inference.

API: unknown · model loaded: false · device: N/A