Digital Media Distribution
Within the School of Digital Media Technology there is focus on the distribution of media across all existing infrastructures and the evolution of new services across these networks.
The broadcast and film industries are continually developing new ways of allowing (and encouraging) the user to interact with content. The research in the school is to propose and develop content creation methods, interactivity and distribution for next generation media formats.
Research activities in the school are focussing on personalisation, interactivity and also the low latency delivery of media over the internet.
An important aspect of this work is that the school works closely with leading industry practitioners, network developers and operators so that (i) the research is industry-centric with immediate relevance; (ii) maintaining the school's profile with industry for the benefit of continued research and also for the research work to augment the school's taught programme content.
The school is rapidly developing its delivery of broadcast technologies programmes with the recent introduction of the BSc Digital Broadcast Technology and MSc Digital Broadcast Engineering. The postgraduate programme is delivered in conjunction with the BBC Academy and delivered for both BBC Graduate Trainees and Birmingham City University-recruited students.
Research in this area is driven by the evolving media industry requirements for processing and networked distribution of media, the increasing requirements for personalised and interactive media, and the requirement for low latency media delivery and interaction.
Two projects are currently being addressed:
Personalised Interactive Online Video
This project addresses the personalisation of in-video content in which video (or audio) objects are selected for individual users and inserted in to the playout in accordance with their profile; different viewers will see different personalised versions of the original programme. The inserted objects are created completely independently of the source media. The source media is tagged to allow and disallow the type of external objects which can be inserted (i.e. allowing the production to maintain an editorial integrity).
The media distribution architecture is essentially a broadband hybrid (c.f. HbbTV). The existing distribution platforms (Cable TV, Satellite, Digital Terrestrial, IPTV, etc) are used with no replacement of installed plant; personalised content is distributed via the web and is united with the source video at the user platform (set top box, PC, mobile, etc).
The scenarios envisaged for personalised interactive media include entertainment shows, personalised advertising and product placement, educational and documentary programmes.
Inserted objects can also be made interactive for end users either individually or as closed user groups.
The concept is further extended:
(i) Media object trading whereby inserted objects for personalisation are brokered in to the source media by a process of auctions and negotiations, for example by agencies representing the providers of the inserted objects. Workflow and metadata descriptions allow the content to determine the limits of externally selected media objects, in this way a producer can protect their content from inappropriate objects being inserted into the programme.
(ii) The Production tools for future programme creators to elegantly create content with interactivity and personalisation.
A placeholder (“#”) is marked during the production of the source video. A selected object (independent of the source video production) is to be transmitted simultaneous to playout to the viewer and will be inserted into the placeholder at the viewer end.
Inserted video objects can be made interactive
The issues for research include:
- Object insertion: image and video processing: i.e. selection and acquisition of objects in a hybrid network; integration, research on standards (e.g. compression and codecs) to be revised and developed for object insertion
- Quality of Experience for the end user: for example ensuring a believable seamless object integration with the source content
- Object selection and brokerage; i.e. the ontologies by which source content and personalisation content is described; user profile acquisition and management
- Object interaction; updating object vectors for user defined placement and motion in the playout scene
- Production tools; i.e. tools for programme production staff who need to define which sections of content allow external content to be added; what restrictions are placed on additional content; interaction capabilities; user-side apps for interaction
The activity currently involves a number of academic, research and industrial organisations across Europe. A number of papers have been published on this research:
- B2B Platform for Media Content Personalisation (B Malheiro, J D Foss, J C Burguillo-Rial); MUSST: Workshop on Multi-User Services for Social TV, EuroITV 2013, Como, Italy; June 2013
- Roadmap for a Comprehensive Evaluation Approach on QoE of Interactive and Personalized TV (J Liebetrau, S Kepplinger, A Cristea, J D Foss);1st International Workshop on Interactive Content Consumption at EuroITV 2013, Como, Italy; June 2013
- Poster: Personalisation of Networked Video (J D Foss, B Malheiro, J C Burguillo-Rial); Futurezone, IBC2012, Amsterdam, Netherlands, September 2012
- Personalisation of Networked Video (J D Foss, B Malheiro, J C Burguillo-Rial); EuroITV, Berlin, Germany, July 2012
- Personalised Placement in Networked Video (J D Foss, B Malheiro, J C Burguillo-Rial); MutliA-Pro, Lyon, France, March 2012
- Negotiation Platform for Personalised Advertising (L V de Sousa, B Malheiro, J D Foss); ECUMICT, Gent, Belgium, March 2012
- Dynamic Personalisation of Media Content (J D Foss, B Malheiro, J C Burguillo-Rial); Semantic Media Adaptation and Personalization (SMAP), Vigo, Spain, December 2011
- A Proposal For Media Brokerage (B.Malheiro, J.D.Foss); ECUMICT, Gent, Belgium, March 2010
Low Latency Media On IP Networks - Flexilink
The volume of audio and video traffic on the web is rapidly expanding and there is an increasing requirement for time critical delivery and response to networked media, especially for user interactive experiences. However there are technical issues raised by the legacy of internet protocol and its routing local and global networks. The University is working with Ninetiles of Cambridge, UK, to realise the prototyping and demonstration of low latency media transmission using Ninetile's Flexilink technology.
The Flexilink technology manages data access to its associated layer 2 data-link protocol to ensure time critical payloads are transmitted in an assured stream alongside traditional best-effort payloads.
DMT-Lab is hosting two PhD projects to progress Flexilink technology:
(i) QoE in a converged time-deterministic and best-effort traffic environment using Flexilink. This project investigates the optimisation of input and output buffering in a Flexilink network node.
(ii) A QoS-oriented protocol for real-time interactive media transmission. This project investigates the network management and performance for a multi-hop Flexilink-based network.
Publications for the Flexilink work include:
- Flexilink: An Unified Low Latency Network Architecture For Multichannel Live Audio
Yonghao (Wang, Grant, Foss); AES 133rd Convention, San Francisco, October 2012
- Poster: Flexilink - A New Low-Latency Network for AV (J Grant, K Grant, Yonghao Wang, J D Foss); Futurezone, IBC2013, Amsterdam, Netherlands, September 2013
Prof Cham Athwal - Associate Head of School (Research)
Cham is a Professor of Digital Technology and Head of Research within the DMT school. His research interests cover 3D modelling, image processing, video processing, digital signal processing web technologies and simulation. Currently Cham is supervising eight PhD/MPhil projects covering a range of subjects including digital audio processing, digital image processing and virtual environments. Email: email@example.com Phone: +44 (0)121 331 5458
Dr Ian Williams
Ian is a Senior Lecturer and Subject leader in Image and Video technology within the school of DMT. His research interests are in low-level image processing, feature extraction and image filtering. Ian currently supervises PhD students in both the image, video and signal processing fields. Email: firstname.lastname@example.org Phone: +44 (0)121 331 7416
Jerry is a senior Lecturer and Subject Leader in Digital Media Technology within the DMT school. His research interests include interactive and personalised media on commercial platforms, involving agent-based brokerage; media distribution architectures and virtual environments applied to environmental construction; learning environments; simulations; spatial interactions and communication in virtual worlds. Email: email@example.com Phone: +44 (0)121 331 7416
Dr Münevver Köküer
Münevver is a Research Fellow within the DMT school. Her research interests cover speech and audio pattern processing, data analytics, pattern recognition, machine learning and image processing. Münevver is currently carrying out research on her specialist areas, developing new research projects and grant applications, developing collaborations and partnerships as well as disseminating research. She supervises PhD projects on digital audio processing, digital image processing and virtual environments. Email: firstname.lastname@example.org Phone: +44 0121 331 7410
Qiang is a lecturer and PhD student in Computer Technology and Science at the Tsinghua University in Beijing. He is currently working as a visiting researcher within the school of DMT and his reseatch interests involve image and video processing systems for target recognition. Qiang has participated in several scientific research projects, including the digital TV monitor of moving people, human face recognition, human fingerprint recognition, auto focus of CCD systems, embedded image collection and storage systems, and high-precision TV-tracking control. Email: email@example.com
Alan is currently working towards his PhD alongside being a visiting lecturer within the DMT school. He's research interests include automated compositing in video and video scene estimation using algorithmic processes, for automated colour coherence, variance analysis in scene illumination and improved match moving.
Email: firstname.lastname@example.org Phone: +44 (0)121 331 7416
Xiangyu Zhu is working on transmission and processing modules for the Aubergine low latency IP-based media project.
We have many emerging areas of research in broadcast technology, media distribution, interactive video and video processing which are available for postgraduate study towards MPhil and PhD awards. In conjunction with projects listed above a number of opportunities are available. For more information on any of these topics please contact Cham Athwal, Jeremy Foss or Yonghao Wang
MSc Digital Broadcast Technology
From September 2014 the school will be delivering this new Masters programme in association with the BBC. The programme requires students to undertake masters projects for their dissertation, and these projects will be closely linked with the growing range of media distribution research activities in the school.