Project Details
Projekt
EVOLVE-QoE - Ecological Validity Evaluation of Interactive Virtual Environments: A QoE Framework for Audiovisual Scenes
Subject Area
Acoustics
Term
since 2020
Project identifier
Deutsche Forschungsgemeinschaft (DFG) - Project number 444832250
The EVOLVE-QoE project builds on the QoEVAVE project from AUDICTIVE Phase 1, aiming to bridge the knowledge gap regarding suitable audiovisual scene characteristics in Interactive Virtual Environments (IVEs). In particular, the aim is to ultimately identify where Interactive Real Environments (IREs), i.e., the real world, and IVEs differ from each other, and which of the respective characteristics induce changes in auditory cognition in complex, real-life-like audiovisual scenes. EVOLVE-QoE adopts a systematic approach to annotate, analyze, and evaluate audiovisual scenes. It employs both top-down human-driven approaches and bottom-up instrumental approaches. The top-down approach captures exploration patterns through 3-DoF to 6-DoF trajectories, eye-tracking, and engagement, providing human-generated annotations based on cognitive processes. The bottom-up approach investigates audiovisual scene analysis, complexity models, and instrumental descriptors. Drawing inspiration from soundscape and visual scene analysis communities, EVOLVE-QoE investigates methods to evaluate the ecological validity of IVE. The project aims to provide suitable schema for characterizing IVEs within a comprehensive methodological framework for QoE assessment. The main objectives of EVOLVE-QoE are i) to extend the IVE scene database (https://qoevave.github.io/database/) with richer 3-DoF and 6-DoF-explorable VR scenes, including scenes with real-life counterparts for ecological validity investigation; ii) to annotate scenes with human-generated labels addressing audiovisual cognition and exploration patterns; iii) to investigate the suitability of instrumental methods for scene characterization and cross-analyze with human-based annotations; iv) to systematically evaluate ecological validity achieved with IVEs compared to IREs using the best subset of human-based and instrumental methods; and v) to integrate all databases, tools, and results into the QoEVAVE/EVOLVE-QoE conceptual and data framework, promoting open science practices. In summary, the EVOLVE-QoE project aims to significantly advance the understanding of audiovisual scene characteristics in IVE research and contribute to the development of a comprehensive framework for QoE evaluation, addressing the AUDICTIVE priorities of IVEs (b), auditory cognition (a), and quality evaluation methods (c).
DFG Programme
Priority Programmes
