Keyword Spotting Using Deep Learning

This repository contains a Streamlit application for keyword spotting using deep learning. The application recognizes specific keywords from audio inputs, utilizing a pre-trained model that we trained using TensorFlow to make predictions from recorded or uploaded audio files.

You can use this model through this streamlit app https://speechkeywordspotting-lxtnkkvezjmbhsxunkk72a.streamlit.app/ or can run it locally following the steps given below.

Features

• Audio Input Options: Users can either upload a .wav audio file or record their voice directly within the app.
• Prediction Visualization: The app displays the predicted keyword probabilities, the audio waveform, and the spectrogram of the audio signal.
• Model: Uses a TensorFlow model that we trained to recognize keywords including "down," "go," "left," "no," "right," "stop," "up," and "yes."

Installation

1. Create and activate a virtual environment (optional but recommended):

   python -m venv venv
   source venv/bin/activate  # On Windows, use `venv\Scripts\activate`

2. Install the required packages:

   pip install -r requirements.txt

3. Run the Streamlit application:

   streamlit run app.py

4. Open your web browser and go to http://localhost:8501.

Usage

Choose between uploading an audio file or recording your voice:

• Upload Audio: Click on "Upload Audio" in the sidebar and upload a .wav file.
• Record Audio: Click on "Record Audio" in the sidebar and record your voice.

Once the audio is processed, view the following results:

• Recorded Audio: Listen to the audio you provided.
• Prediction Results: View a bar chart of the predicted keyword probabilities.
• Audio Waveform: See the waveform of the audio signal.
• Spectrogram: Visualize the spectrogram of the audio signal.