Mittwoch, 11. November 2026 · 14:35–15:00 Uhr · Raum 1
Session 5: Next Generation Audio, Accessibility & Media Quality
Abstract
We introduce a novel deep learning–based audio-visual quality (AVQ) prediction model that leverages internal features from state-of-the-art unimodal predictors. Unlike prior approaches that rely on simple fusion strategies, our model employs a hybrid representation that combines learned Generative Machine Listener (GML) audio features with handcrafted Video Multimethod Assessment Fusion (VMAF) video features. Attention mechanisms capture cross-modal interactions and intra-modal relationships, yielding context aware quality representations. A modality relevance estimator quantifies each modality’s contribution per content, potentially enabling adaptive bitrate allocation. Experiments demonstrate improved AVQ prediction accuracy and robustness across diverse content types.
Referentenprofil Ina Salaj – Dolby Germany GmbH
Ina Salaj is a Software Tester at Dolby Laboratories, where she works on testing and validating audio software and systems within complex development pipelines. She joined Dolby in 2023 after a background in AI development and applied research. She has an academic background in Geoinformatics Engineering and Computer Science, with a focus on audio quality and AI-based methods. Her research paper on audio-visual quality prediction was accepted and presented at ICASSP 2026.
Jetzt Ticket sichern: Tickets für Mitglieder · Tickets für Externe
