Speech Disorder Detection And Correction System

Overview

This pipeline is designed to help children improve their pronunciation and vocabulary through interactive learning activities. The game presents two modes of play:

1. Object Identification Mode: Displays an image of a regular object. The child has to speak the name of the object when prompted.
2. Sentence Reading Mode: Displays a short story, sentence by sentence, and the child reads aloud. Feedback is provided on mispronunciations or stammering.

The game dynamically adjusts the experience to ensure children learn effectively, with real-time feedback and guidance.

---

Features

Object Identification Mode

• Displays an image of a common object (e.g., apple, car, dog).
• A timer counts down, after which the microphone is enabled.
• The child speaks the name of the object aloud.
• Correct Response: Moves to the next image.
• Incorrect Response: The game repeats the correct name and prompts the child to try again.

Sentence Reading Mode

• Displays a short, moral-based story, sentence by sentence.
• The child reads each sentence aloud.
• Feedback Mechanism:
  • Detects stammering or mispronunciation.
  • Repeats the sentence as audio output.
  • The child must re-read the current sentence until pronounced correctly.
• Includes a Hint Button: Pronounces the current sentence word by word to assist the child.

---

Workflow

1. Image-Based Interaction:
   • Timer counts down.
   • Speech is captured through a microphone.
   • Speech-to-text conversion occurs.
   • Text is matched against the object name for validation.
2. Story-Based Interaction:
   • Sentence from the story is displayed.
   • Child’s speech is analyzed using speech-to-text.
   • Validation for correct pronunciation.
   • Cursor updates to the next sentence upon successful pronunciation.
   • Feedback provided for errors, prompting the child to repeat.
3. Feedback Loop:
   • Visual and audio feedback ensures engagement.
   • Repeat mechanisms reinforce learning.

---

Technologies Used

• Streamlit: For developing the user interface.
• Speech-to-Text API: Converts the child’s speech into text for analysis.
• Python Libraries:
  • speech_recognition: For capturing and processing audio.
  • Pillow: For displaying images.
  • pygame: For playing audio feedback.
• Natural Language Processing (NLP):
  • Used for analyzing speech and providing feedback.

---

Installation

1. Clone the repository:

   git clone https://github.com/ARYANSINGH0611/PronouncePerfect.git

2. Install dependencies:

   pip install -r requirements.txt

3. Run the application:

   streamlit run app.py

---

Usage

1. Launch the game.
2. Select the desired mode: Object Identification or Sentence Reading.
3. Follow the on-screen instructions to complete the activity.
4. Use the feedback to improve pronunciation and vocabulary.

---

Future Improvements

• Expand the object image database with more diverse objects.
• Add difficulty levels for story complexity and vocabulary.
• Incorporate multilingual support for non-English learners.
• Include gamified rewards to enhance engagement.

---

Contribution

Contributions are welcome! Please open an issue or submit a pull request for any feature suggestions or bug fixes.