In a telephone link, the voice timbre is impaired by spectral distortions generated by the analog parts of the link. We first evaluate from a perceptual point of view an equalization method consisting in matching the long term spectrum of the processed signal to a reference spectrum. This evaluation shows a satisfying restoration of the timbre for most speakers. For some speakers however, a noticeable spectral distortion remains. That is why we propose a multi-referenced equalizer, based on a classification of speakers and using a different reference spectrum for each class. This leads to a decrease of the spectral distortion and, as a consequence, to a significant improvement of the timbre correction.
The rapid growth of wireless mobile communications raises the critical issue of meeting user expectations and requirements. This paper deals with the subjective assessment of the perceived quality of video calling services delivered on a real LTE/4G network. The main goal is to determine the list of parameters that are predominantly impacting the Quality of Experience (QoE). The influencing parameters considered here include the video sequence content, coding technology and both core and radio access networks. Following the ITU-T Recommendations, subjective quality assessment experiments were conducted. Subjects were shown different representative sequences of video-calling services transmitted using different network conditions. The outcomes of these experiments show that the QoE is mainly affected by packets loss, jitter. However the impact of these parameters, which vary over time and in case of user mobility, seems to be different from that observed with simulated networks.
This paper analyzes the actual usage of Mean Opinion Scores (MOS) in the telecommunication industry and looks back to the tremendous evolutions that has encountered this sector since subjective test methodologies were standardized by ITU-T. It is pointed out that MOS have severe drawbacks and that test methods should evolve by taking into account valuable inputs from connected research fields such as the psychology of emotions, cognitive science, usability engineering and marketing research.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.