I'm a software engineer passionate about building audio applications, combining my love for music and programming. Although this journey can be quite challenging, I’m excited about the profound impact music can have on people's well-being, and not to mention, the potential social benefits.
My professional interests involve C, C++, and Python, primarily working in Linux environments. I am acquainted with Data Structures, Algorithms, Digital Signal Processing, and Machine Learning. Having contributed to various stages of development, I have experience with UML design, Testing, Continuous Integration, Dockerizing, and using various tools to improve workflows. Currently, I’m exploring JUCE and Cross-Platform Development, and in the future, I’m looking to expand into ONNX and RTOS.
🚀 If you're interested in collaborating or have a potential opportunity in mind, feel free to reach out. If you're just browsing and enjoy my work, I’d love to hear your thoughts or see you star my projects below! 👇
Interactive Audio Visualizer: Bringing Sound to Life with Interactive Real-Time Audio-Visual Experience
This project develops an innovative interactive tool designed to offer a real-time audiovisual experience for users. The envisioned objective of IAV is to provide individuals with limited mobility or minimal musical background an accessible way to engage with positive musical stimuli, using simple actions like hand movements, to inspire and motivate them to participate in creative expression.
Pi-Loop-Console: A linux based audio looper written in C++.
Pi-Loop is a real-time audio looper application that enables users to create interactive musical sessions. One may use an external audio interface to connect a microphone and an instrument or to use the integrated sound card to test it.
My diploma thesis: Audio signal classification using deep learning algorithms.
In this thesis we compared the performance of multiple feature parameters for environmental sound classification problems by developing multiple evaluating models. Specifically, as audio representation of two different datasets, we used raw waveforms, log-mel spectrograms and short-time Fourier transforms. Finally we set four different experiments and each one of them was divided in two discrete audio representation modes. For their evaluation and also for comparability purposes we developed hybrid CNN models. Along with comparing each mode within each experiment, we also compared the performances achieved by using each different dataset through inspecting and examining the factors of structure, the technical features and various prospects of the initial data distribution, respectively for each dataset. The nature of this research additionally enabled us to seek for potential environmental class-conditional audio features.
The SMOD Project: a complex visualization of sound using an artistic approach
A simple presentation of an audio visualizer with artistic design, utilizing diverse components and a MATLAB script to generate real-time analog and digital visuals synced with audio playback.
