This repository contains experimental work on developing an AI-powered Natural Language Processing (NLP) conversational interface with voice recognition capabilities for a digital human.
- Speech-to-text conversion
- Natural language understanding
- Conversational interface
- Text processing and punctuation restoration
- Diarization
- Multi-lingual support
- Integration with audio2face-3d
- Python
- Vosk speech recognition models
- Custom NLP components
This is an experiment / work in progress but intended to be part of a digital human conversational interface and part of an agentic system that will use smart context management along with a conversational interface to enable fluid conversation while minimizing token usage or compute required.