Hi there I'm Thodoris, from Athens, Greece. I am a Principal Researcher of Multimodal Machine Learning. I work on ideas for using machine learning and signal processing to analyze sounds, images, and videos.

Open-source projects

I've worked on several Python open-source projects, such as

• pyAudioAnalysis (A library for audio feature extraction and recognition),
• multimodal_movie_analysis (a library for analyzing the raw content of movies, i.e. video, audio and text)
• amvoc (a tool for analysis of mouse vocal communication),
• paura (a python library for real-time recording and analysis of sounds) and
• deep_audio_features (a Pytorch implementation of deep audio embedding calculation).

Educational material and articles

I also like to create educational content for multimodal ML issues such as:

• This article on Intro to basic audio handling and this accompanying github repo.
• Or this intro article in Audio recognition using Python with this jupyter notebook material.

Future goals and why to support me

I am currently working on extending the core functionalities of the audio libraries but also working on existing and new libraries that demonstrate how ML can be used to achieve impressive results in application domains such as movie recommendation and speech analytics.. Sponsorship will really make it more sustainable for me to continue working on these open source projects, maintain them and support them through computational (mostly) resources.