IT pet projects
IT pet projects
Flask app: Game of Life
This is a training project in the Flask learning course. The project realizes the Game of Life (GoF) in a browser. Two versions are realized: (i) life in a box and (ii) life on a thorus. They differ in how the number of the cell's neighbors is calculated. The app is deployed on Heroku.com using a Docker container. You are welcome to test it using the link above.
Flask app: Sound classification
This project aims to develop a machine learning (ML) model to classify sounds produced by three types of vehicles in the city streets: buses, cars, and trams. For this reason, the dataset of such sounds was collected. For audio signal processing and machine learning classification, we use Python packages librosa and sklearn, respectively. The model showed a decent accuracy of about 90% for sound classification. Therefore, it was decided to write a simple Flask app that classifies the user-uploaded audio files.
The logic of sound recognition is developed using the trained kNN model using the data set of 380/class recordings. The app is deployed on AWS lightsail container service using Docker container. You are welcome to test it using the link above. Be ready to upload approximately 5-second audio in wav format.