About me
Researcher at the Department of Multimedia Systems, Gdańsk University of Technology, with a focus on audio machine learning, psychoacoustics, signal processing, automatic speech recognition, and deepfake audio detection. His work sits at the intersection of immersive audio and AI, exploring how neural network-based approaches can be applied to spatial audio processing and audio forensics. At AES Europe 2026, he is presenting "The Ambisonic Denoising Paradox: U-Net Processing Degrades ASR Transcription Quality for Medical Speech" (co-authored with Bartłomiej Mróz), which introduces the Medical Immersive Audio Corpus (MIAC) and reveals a counterintuitive finding: U-Net denoising of third-order Ambisonic recordings can degrade - rather than improve ASR performance for most models. Szymon is a longtime AES member and a Silver Medal winner of the AES Student Design Competition (AES Berlin Convention, 2016) for a rapid prototyping system for acoustic treatment design.