arthur-monnet / ia_noise_recognition

Sound Recognition AI

This project is a simple sound recognition AI system built using Keras and TensorFlow. It can recognize three different sounds: the hoot of an owl, a hunter's shot, and the chirping of a robin.

Prerequisites

Before running the program, make sure you have Python 3.10 installed on your machine. You will also need to install the following libraries:

• numpy
• tensorflow
• pandas
• librosa
• keras
• pyaudio

You can install the required libraries by running the following command in your terminal:

pip install -r requirements.txt

Getting Started

To start, you will need to have sound files of the three sounds you want to recognize. Place the sound files in the data directory under a subdirectory with the name of the sound. For example, if you want to recognize the hoot of an owl, create a subdirectory called owl_hoot in the data directory and place owl hoot sound files in it.

Data Preprocessing

Before training the AI system, you need to preprocess the sound files and create the necessary data files. To do this, run the data_preprocessing.py file:

python data_preprocessing.py

The preprocessed data will be saved in the data_preprocessed directory.

Model Training

Once the data preprocessing is complete, you can train the AI system. To do this, run the model_training.py file:

python model_training.py

The trained model will be saved in the saved_models directory.

Real-time Sound Recognition

To recognize sounds in real-time using the microphone, run the real_time_sound_recognition.py file:

python real_time_sound_recognition.py

This will initialize the microphone and start recognizing sounds. The recognized sound class will be printed to the console.

Testing the Model

To test the accuracy of the trained AI system, run the model_testing.py file:

python model_testing.py

This will evaluate the model on the test data and print a classification report.

Making Predictions

You can also use the trained model to make predictions on new sound files. To do this, run the prediction.py file:

python prediction.py path/to/sound/file.wav

Replace path/to/sound/file.wav with the path to the sound file you want to predict the class of. The predicted sound class will be printed to the console.