Anel Music's Projects
This project was created more or less out of boredom on a Sunday afternoon. The goal is to read a video stream or play a recorded video thereby classifying the movements of the user and play a corresponding piece of music according to the recognized musical instrument (guitar, piano, violin).
The goal of this project is to illustrate how easy it is to perform image segmentation today. With relatively little effort it is possible to achieve acceptable results. Here we use a FCN8 model with a pretrained VGG16 backbone.
Example Repo for the Udemy Course "Deployment of Machine Learning Models"
In this notebook, we describe the entire process from data preparation/preprocessing to training the network using transfer learning and fine-tuning. Finally we will analyze the classification performance and present the results.
Template for the app using FastAPI/MySQL, that Includes user model & auth functionality, and no ORM, minimal set.
This project is used to illustrate the convenient usage of FastAPI.Here, we will use a local (in memory) database to store location objects. Using a REST API we can access the database information and create and store location objects.
App that lets you 'tune' your activity diagrams – because who wouldn't want to outsmart those who think they can judge your software engineering skills based on diagrams alone?
In this project, an interactive web application is deployed that allows the user to select a dataset and a classifier. The parameters of the classifiers can be changed freely. You can run the app on your localhost or use the deployed webapp
Minikube-powered Kubernetes cluster setup showcasing internal and external pod communication. Demonstrates FastAPI and NGINX deployments utilizing LoadBalancer and ClusterIP services. Internal pod interaction via ClusterIP and external access through Restful API via LoadBalancer.
Minikube-powered Kubernetes cluster setup showcasing internal and external pod communication. Demonstrates MongoDB and Mongo Express deployments utilizing LoadBalancer and ClusterIP services. Internal pod interaction via ClusterIP and external access to MongoDB provided through a user-friendly interface via LoadBalancer.
To be able to experiment, iterate, debug and deploy quickly and easily when working in a team, it is necessary to create a dedicated environment. It is especially important to be able to reproduce the state of a project (code, data, model). This is the core function of the so-called machine learning devops (short MLOps). This project can serve as a kind of template for those interested in setting up such an environment and is intended to show the main advantages.
Train a Language Model capable of explaining Python code and providing explanations with a fun twist inspired by Hollywood mafia classics such as “The Godfather”, “Casino”, and “Goodfellas.
This project is a modern web application utilizing FastAPI for the backend API and PostgreSQL for the database.
This is an implementation of Particle Filter Localization in Python using the Pygame library. The main goal of this project is to demonstrate how a robot can localize itself within an environment based on sensor measurements.
This visualization aims to show the functionality of commonly used search algorithms step by step. The user interface allows you to create walls, set a start and a target position and choose any search algorithm.
Using voice commands to control DC and Servo motors
The goal of this project is to illustrate how easy it is to perform image segmentation with Pytorch. With relatively little effort it is possible to achieve acceptable results. Here we use a DeeplabV3 model with a ResNet101 backbone.
Concise, consistent, and legible badges in SVG and raster format