Blind quality algorithm to analyze streaming video contents in 5g networks

López Velasco, Juan Pedro; Authier, Anton; Jiménez Bermejo, David; Menéndez García, José Manuel y Sánchez Almodóvar, Nuria (2017). Blind quality algorithm to analyze streaming video contents in 5g networks. En: "International Conferences WWW / Internet 2017 and Applied Computing 2017", 18 – 20 October 2017, Algarve, Portugal.

Descripción

Título: Blind quality algorithm to analyze streaming video contents in 5g networks
Autor/es:
  • López Velasco, Juan Pedro
  • Authier, Anton
  • Jiménez Bermejo, David
  • Menéndez García, José Manuel
  • Sánchez Almodóvar, Nuria
Tipo de Documento: Ponencia en Congreso o Jornada (Artículo)
Título del Evento: International Conferences WWW / Internet 2017 and Applied Computing 2017
Fechas del Evento: 18 – 20 October 2017
Lugar del Evento: Algarve, Portugal
Título del Libro: International Conferences WWW / Internet 2017 and Applied Computing 2017
Fecha: 20 Octubre 2017
Materias:
Palabras Clave Informales: Streaming, 5G networks, Video Quality, Artefacts, Metrics, QoE
Escuela: E.T.S.I. Telecomunicación (UPM)
Departamento: Señales, Sistemas y Radiocomunicaciones
Grupo Investigación UPM: Aplicación de Telecomunicaciones Visuales G@TV
Licencias Creative Commons: Reconocimiento

Texto completo

[img]
Vista Previa
PDF (Document Portable Format) - Se necesita un visor de ficheros PDF, como GSview, Xpdf o Adobe Acrobat Reader
Descargar (593kB) | Vista Previa

Resumen

Internet traffic is growing by an average of 40% a year and video streaming stands for 50% of the traffic. The expansion of new applications and services for video distribution makes users demand video quality and faster loads for contents with higher resolutions. For that reason, it is necessary to improve detection systems with check the received contents and insure comfortable use. The wide variety of technologies for streaming and encoding systems make necessary the implementation of blind quality models that detect errors independently of the transmission source, focused on analyzing the decoded image as consumed by the final user. Additionally, the absence of a video reference in streaming systems oblige the creation of No-Reference quality metrics, because the content is not available to compare the video received with the original source. For that reason, this paper proposes a novel algorithm based on the detection transmission artefacts using visual VQA techniques in order to improve users' Quality of Experience (QoE). The algorithm is based on the detection of main streaming visual errors and artefacts, such as color degradation, frozen frames, complete absence of video or packet losses. The algorithm is included in a quality probe application that is part of a 5G transmission network system for streaming high quality video. The algorithm analyzes the video quality for detecting streaming errors and the application warns the broadcaster for the retransmission of content if necessary. For the evaluation of the algorithm, a database of videos containing different types of errors was used in order to simulate streaming conditions and approach reality. The development of this application is necessary for the introduction of next generation 5G telecommunications networks and the consequent new paradigm of video broadcasting.

Proyectos asociados

TipoCódigoAcrónimoResponsableTítulo
Horizonte 20206715985G-CrosshaulUniversidad Carlos IIIFuture Internet: Software, hardware, Infrastructures, technologies and services

Más información

ID de Registro: 48650
Identificador DC: http://oa.upm.es/48650/
Identificador OAI: oai:oa.upm.es:48650
Depositado por: Juan Pedro López Velasco
Depositado el: 01 Dic 2017 10:13
Ultima Modificación: 01 Dic 2017 10:13
  • Open Access
  • Open Access
  • Sherpa-Romeo
    Compruebe si la revista anglosajona en la que ha publicado un artículo permite también su publicación en abierto.
  • Dulcinea
    Compruebe si la revista española en la que ha publicado un artículo permite también su publicación en abierto.
  • Recolecta
  • e-ciencia
  • Observatorio I+D+i UPM
  • OpenCourseWare UPM