Haupt-Navigation

Inhalt des Dokuments

Dr.-Ing. Sebastian Arndt

: © Copyright??

: © Copyright??

Research Field:

Quality

Research Topics:

Neurotechnology

Current Project

PflegeTab

Past Projects

Biography:
Sebastian Arndt is a researcher at the Quality and Usability Lab of the Telekom Innovation Laboratories, TU Berlin. He studied Computer Science at Technische Universität Berlin and received his diploma in 2010. He received his doctoral degree (Dr.-Ing.) in 2015 with the thesis title 'Neural Correlates of Quality During Perception of Audiovisual Stimuli'. His current research focus is on physiological changes during the perception of audiovisual quality. 2012 he was a visiting researcher at MuSAE Lab (Montreal, Canada) in order to work on multimodal neural correlates for synthesized speech.
Address:

Quality and Usability Lab
Deutsche Telekom Laboratories
TU Berlin
Ernst-Reuter-Platz 7
D-10587 Berlin, Germany
Telefon:
+49 30 8353 58328
+49 1709147458 E-Mail: sebastian.arndt@telekom.de

Publications

Neural correlates of Affective States and Speech Perception: Towards Quality-of- Experience Measurement of Reverberant Speech
Zitatschlüssel	antons2013e
Autor	Antons, Jan-Niklas and Laghari, K. and Schleicher, Robert and Arndt, Sebastian and Falk, Tiago H. and Möller, Sebastian
Buchtitel	Proceedings of the International Conference on Acoustics (AIA-DAGA 2013)
Seiten	1–4
Jahr	2013
ISBN	9783939296058
Ort	Merano, Italy
Adresse	Berlin, Germany
Notiz	Electronic/online
Verlag	German Acoustical Society (DEGA)
Serie	AIA-DAGA
Wie herausgegeben	Full
Zusammenfassung	Speech quality assessment usually depends on subjective judgments after listening to test stimuli. The obtained subjective quality indices are valid and reliable but provide little insight into the underlying perceptual process. Since reverberation is known to influence perceived speech quality and intelligibility, e.g. in conference calls while using a loudspeaker, we analyzed the performance of electroencephalography (EEG), which measures brain activity at the cortical level, for indicating speech stimuli with a high reverberation time as degradation. We collected a database of 22 subjects to test the ability of utilizing EEGdata - especially parameters of event-related-potentials (ERP) - to identify the processing of a stimulus with high reverberation times. The reported preliminary findings provide promising insights: indirect measurements of perceived stimulus quality - without asking for the subjects' opinion - are sensitive to reverberation levels. Correlations between physiological parameters and stimulus features showed that quality degradations can be monitored in conscious stages of stimulus processing. We show that the analysis of ERP is in general a useful and valid tool in quality research. In the case of reverberation, this can actually lead to the indirect measurement of perceived quality with respect to changes of room acoustics.