AsQM: Audio streaming Quality Metric based on Network Impairments and User Preferences
There are many users of audio streaming services because of the proliferation of cloud-based audio streaming services for different content. The complex networks that support these services do not always guarantee an acceptable quality on the end-user side. In this paper, the impact of temporal interruptions on the reproduction of audio streaming and the users preference in relation to audio contents are studied. In order to determine the key parameters in the audio streaming service, subjective tests were conducted, and their results show that users Quality-of-Experience (QoE) is highly correlated with the following application parameters, the number of temporal interruptions or stalls, its frequency and length, and the temporal location in which they occur. However, most important, experimental results demonstrated that users preference for audio content plays an important role in users QoE. Thus, a Preference Factor (PF) function is defined and considered in the formulation of the proposed metric named Audio streaming Quality Metric (AsQM). Considering that multimedia service providers are based on web servers, a framework to obtain user information is proposed. Furthermore, results show that the AsQM implemented in the audio player of an end users device presents a low impact on energy, processing and memory consumption.
Code (0)
등록된 구현이 없습니다.
Methods 이 논문이 사용한 방법론
Similar Papers 제목 키워드 기반
VISTA: A Benchmark for Real-Time Video Streaming under Network Impairments in Surgical Teleoperation
Real-time video streaming is crucial in surgical teleoperation, yet reproducible evaluation under realistic network impairments remains limited. This paper presents VISTA, a benchmark designed to study how impairments al…
Multimodal Confidence Modeling in Audio-Visual Quality Assessment
Audio-visual quality assessment (AVQA) is essential for streaming, teleconferencing, and immersive media. In realistic streaming scenarios, distortions are often asymmetric, where one modality may be severely degraded wh…
Breaking the Quality--Intelligibility Trade-off in Streaming Target Speaker Extraction via Deep-Feature-Anchored Preference Optimization
Generative streaming models for Target Speaker Extraction (TSE) commonly exhibit a quality--intelligibility trade-off, wherein naive optimization for perceptual audio quality tends to degrade speech intelligibility, and …
Impairments are Clustered in Latents of Deep Neural Network-based Speech Quality Models
In this article, we provide an experimental observation: Deep neural network (DNN) based speech quality assessment (SQA) models have inherent latent representations where many types of impairments are clustered. While DN…
ClassificationSupervised Classifiers for Audio Impairments with Noisy Labels
Voice-over-Internet-Protocol (VoIP) calls are prone to various speech impairments due to environmental and network conditions resulting in bad user experience. A reliable audio impairment classifier helps to identify the…