This project aims to develop a comprehensive system for streamlining therapy sessions by automating key tasks, allowing therapists to focus on patient interaction.
- Speaker Identification: Accurately identifies therapists and patients within the recorded session audio.
- Speech-to-Text Conversion: Converts spoken dialogue from the therapy session into a text transcript.
- Refined Progressive Note Generation: Creates refined and improved progress notes based on the transcribed text, potentially including sentiment analysis and summarizing key points.
- Suggestion System Integration: Integrates a suggestion system that leverages NLP models and large language models (LLMs) to provide therapists with real-time or post-session prompts, resources, or interventions tailored to the conversation.
- Deep Learning Models: Used for speaker identification.
- Automatic Speech Recognition (ASR) Models: Used for speech-to-text conversion.
- Natural Language Processing (NLP): Techniques for text analysis and summarization.
- Large Language Models (LLMs): For potential suggestion generation and information retrieval.
- Increased Efficiency: Automates transcription, reducing therapists' time spent on note-taking.
- Improved Accuracy: Ensures consistent and accurate capture of session details.
- Enhanced Analysis: Enables advanced note analysis and potential suggestion generation to support therapeutic decisions.
This project has the potential to significantly improve the workflow for therapists, allowing them to dedicate more time to patient care and interaction.
- Python 3.8+
- Pip
- Hugging Face token
- OpenAI API key
- Firebase credentials JSON file
git clone https://github.com/22036435/cogn.io.git
cd therapy-session-analysis