HAN Speech Command Model
Overview
Speech recognition model for humanoid robot voice commands. Optimized for real-time command detection.
Model Type
Conformer-based audio model
Training Data
- Voice command dataset
- Synthetic speech samples
- Noise-augmented audio clips
Input
- Audio waveform (16kHz)
- Mel spectrogram features
Output
- Command classification
- Intent prediction
Framework
PyTorch + Torchaudio
Use Case
- Voice control
- Smart assistant integration
- Human-robot interaction
License
MIT