Tuesday, May 23, 10:30 — 12:00
Jordan Craig (Chair)
P24-01 Usability and Effectiveness of Auditory Sensory Substitution Models for the Visually Impaired
Adam Csapo (Presenting Author), Michal Bujacz (Author), Marcelo Herrera Martinez (Author), Gabriel Ivanica (Author), Maciej Janeczek (Author), Alin Moldoveanu (Author), Simone Spagnol (Author), Runar Unnthorsson (Author), György Wersényi (Author)
This paper focuses on auditory sensory substitution for providing visually impaired users with suitable information in both static scene recognition and dynamic obstacle avoidance. We introduce three different sonification models together with three temporal presentation schemes, i.e., ways of temporally organizing the sonic events in order to provide suitable information. Following an overview of the motivation and challenges behind each of the solutions, we describe their implementation and an evaluation of their relative strengths and weaknesses based on a set of experiments in a virtual environment.
Convention Paper 9801
P24-02 Adaptive Audio Engine for EEG-Based Horror Game
Jordan Craig (Presenting Author)
This paper documents the design and play-testing of a videogame that incorporates electroencephalography (EEG) technology to augment traditional controls. A survival horror game was created using Unity3D. The player navigates the game using conventional keyboard and mouse movement, however, they also wear an Emotiv EPOC headset that transmits their level of calm to the game via OSC. In order to complete the game, the player must remain as calm as possible. An adaptive audio engine was developed to act as an auditory display for this complex parameter in lieu of a distracting visual indicator. Every element of the audio was designed to adapt to the constantly fluctuating value. Procedural audio modules were created in Max, where player EEG data was simultaneously mapped to a myriad of modulators. FMOD Studio was used for non-procedural elements due to its facilitation of real-time control parameters, as well as its integration with Unity3D.
Convention Paper 9802
P24-03 Real-Time Reverb Reduction for Improved Automatic Speech Recognition in Far-Field
Adam Kupryjanow (Presenting Author), Lukasz Kurylo (Author), Piotr Lasota (Author), Przemyslaw Maziewski (Author)
In the paper, methods of real-time reverb reduction based on Generalized Weighted Prediction Error (GWPE) were presented. It was shown that usage of the proposed audio processing routines highly improve the accuracy of Automatic Speech Recognition (ASR) system namely word error rates (WERs) are reduced 11.36% when the user stands 5 meters from the microphone array. The obtained results are close to the ones that are achieved by the offline GWPE implementation (12.06%). Thanks to optimizations and parameters tuning, computational complexity of the proposed realization of GWPE was highly reduced and it achieves RTFs lower than 1.0 (computation time is shorter than signal duration) when using one core of CPU.
Convention Paper 9803