|
|
--- |
|
|
title: CASL 2 - Speech Therapy Assessment Tool |
|
|
emoji: 🎤 |
|
|
colorFrom: blue |
|
|
colorTo: purple |
|
|
sdk: gradio |
|
|
sdk_version: "3.50.0" |
|
|
app_file: app.py |
|
|
pinned: false |
|
|
--- |
|
|
|
|
|
# CASL 2 - Speech Therapy Assessment Tool |
|
|
|
|
|
An interactive tool for speech therapists to assess and treat speech disorders. CASL 2 combines professional speech therapy assessment with AI-powered feedback, making it easier for therapists to conduct evaluations and track progress. |
|
|
|
|
|
## Features |
|
|
|
|
|
### 1. Conversation Assistant |
|
|
- Natural voice-based conversation with an AI assistant |
|
|
- Get information about speech therapy techniques and assessment methods |
|
|
- Guidance on using the assessment tools |
|
|
- Full conversation history tracking |
|
|
|
|
|
### 2. Articulation Assessment |
|
|
- Evaluate speech sound production in various positions (initial, medial, final) |
|
|
- Visual prompts with images for each target word |
|
|
- Real-time recording and transcription |
|
|
- Automatic analysis of pronunciation accuracy |
|
|
- AI-powered feedback for each word |
|
|
- Progress tracking through the assessment |
|
|
|
|
|
### 3. Language Assessment |
|
|
- Evaluate receptive and expressive language skills |
|
|
- Test vocabulary, following directions, and sentence formation |
|
|
- Record responses and get professional feedback |
|
|
- Structured assessment format with varied task types |
|
|
|
|
|
## Setup |
|
|
|
|
|
### Hugging Face API Token |
|
|
|
|
|
This app requires a Hugging Face API token to access LLM services. To set it up: |
|
|
|
|
|
1. Create an account on [Hugging Face](https://huggingface.co/) |
|
|
2. Generate an API token in your [settings page](https://huggingface.co/settings/tokens) |
|
|
3. When deploying to Hugging Face Spaces: |
|
|
- Go to Settings > Repository Secrets |
|
|
- Add a secret with the key `HF_API_TOKEN` and your token as the value |
|
|
|
|
|
### Local Development |
|
|
|
|
|
To run this application locally: |
|
|
|
|
|
1. Install the required dependencies: |
|
|
|
|
|
```bash |
|
|
pip install -r requirements.txt |
|
|
``` |
|
|
|
|
|
2. Set your Hugging Face API token: |
|
|
|
|
|
```bash |
|
|
export HF_API_TOKEN="your-token-here" |
|
|
``` |
|
|
|
|
|
3. Run the application: |
|
|
|
|
|
```bash |
|
|
python app.py |
|
|
``` |
|
|
|
|
|
4. Open the URL shown in the terminal (typically http://127.0.0.1:7860) |
|
|
|
|
|
## How to Use |
|
|
|
|
|
### Conversation Mode |
|
|
1. Select the "Conversation Assistant" tab |
|
|
2. Click "Start Conversation" to begin |
|
|
3. Record your speech using the microphone |
|
|
4. The AI will respond with voice and text |
|
|
5. Continue the conversation naturally |
|
|
|
|
|
### Articulation Assessment |
|
|
1. Select the "Articulation Assessment" tab |
|
|
2. Click "Start Assessment" to begin |
|
|
3. View the current word and associated image |
|
|
4. Ask the patient to pronounce the displayed word |
|
|
5. Record their response using the microphone |
|
|
6. Review the AI analysis and feedback |
|
|
7. Use the navigation buttons to move between words |
|
|
8. Complete all words in the assessment |
|
|
|
|
|
### Language Assessment |
|
|
1. Select the "Language Assessment" tab |
|
|
2. Click "Start Assessment" to begin |
|
|
3. Present the current task to the patient |
|
|
4. Record their response using the microphone |
|
|
5. Review the AI analysis and feedback |
|
|
6. Use the navigation buttons to move between tasks |
|
|
7. Complete all tasks in the assessment |
|
|
|
|
|
## Technical Details |
|
|
|
|
|
This app combines several technologies: |
|
|
- Gradio's UI components and audio capabilities |
|
|
- Google's Speech Recognition API for transcription |
|
|
- Hugging Face's LLM API (Llama-2-7b-chat-hf model) |
|
|
- gTTS (Google Text-to-Speech) for voice synthesis |
|
|
- Simple speech analysis for assessment purposes |
|
|
- Modern web interface with responsive design |
|
|
|
|
|
## For Therapists |
|
|
|
|
|
CASL 2 is designed to supplement professional speech therapy assessment, not replace it. The tool provides an engaging interface for patient interaction and helps therapists with: |
|
|
|
|
|
1. Recording and documenting patient responses |
|
|
2. Initial screening of speech patterns |
|
|
3. Tracking progress over time |
|
|
4. Engaging patients with interactive exercises |
|
|
5. Providing consistent, supportive feedback |
|
|
|
|
|
Always combine the tool's analysis with your professional judgment. |
