API Build-data JSON Resources
Theme

Choose how MSRBot.io looks on this device.

Auto (match system) Light mode Dark mode

Preference is stored in this browser only.

SMPTE Motion Imaging Journal ( Volume: 129, Issue: 2, March 2020)
[ACTIVE]

AI IN PRODUCTION: VIDEO ANALYSIS AND MACHINE LEARNING FOR EXPANDED LIVE EVENTS COVERAGE

Metadata

Publisher
SMPTE — White Plains, NY, USA
Doc Type
Journal Article
Content Type
Original Research
Abbreviated Title
SMPTE Mot. Imag. J
Volume
129, No. 2, pp. 36–45
Abstract
As with many industries, TV and video production is likely to be transformed by artificial intelligence (AI) and machine learning (ML), with software and algorithms assisting production tasks that, conventionally, could only be carried out by people. Expanded coverage of a diverse range of live events is particularly constrained by the relative scarcity of skilled people, and it is a strong use case for AI-based automation. This article describes the recent research conducted by the British Broadcasting Corporation (BBC) on the potential production benefits of AI algorithms, using visual analysis and other techniques. Rigging small, static ultrahigh-definition (UHD) cameras, we have enabled a one-person crew to crop UHD footage in multiple ways and cut between the resulting shots, effectively creating multicamera HD coverage of events that cannot accommodate a camera crew. By working with programmakers to develop simple deterministic rules and, increasingly, training systems using advanced video analysis, we are developing a system of algorithms to automatically frame, sequence, and select shots, and construct acceptable multicamera coverage of previously untelevised types of events.
Publication Date
2020-03-01
DOI
10.5594/JMI.2020.2967204
ISSN
Print: 1545-0279 | Electronic: 2160-2492
Link
https://doi.org/10.5594/JMI.2020.2967204
Author(s)
Craig Wright
bio
Craig Wright is a systems engineer turned research engineer with a background in systems programming, artificial intelligence, and computer vision. Since joining BBC Research and Development in 2016, he has contributed to projects including, the delivery of binaural audio and object-based audiovisual experiences in the browser; and the automatic detection of semantic programme to improve content search. Most recently, he has worked with a team of multidisciplinary engineers to prototype Ed: a system that automatically produces an edited video package from several input video streams captured using unmanned cameras.
Jack Allnutt
bio
Jack Allnutt is a software engineer with an interest in responsible technology, particularly on privacy and data protection. Since joining BBC Research and Development in 2014, he has contributed to projects including the development of object-based media experiences, the use of artificial intelligence (AI) in media production, the use of personal data stores in a public service environment and providing audiences access to innovative and experimental technologies.
Rosie Campbell
bio
Rosie Campbell is a program lead at the Partnership on Artificial Intelligence (PAI), San Francisco, CA, where she leads a research program on responsible publication practices for AI and machine learning (ML). Before joining PAI, she was assistant director of the Center for Human-Compatible AI, an AI safety research group at UC Berkeley, Berkeley, CA, working toward provably beneficial AI. Previously, she was a research engineer at the BBC Research and Development in the Future Experience Technologies Team, where she worked on a variety of experimental technical projects spanning AI, computer vision and graphics, and web-based production interfaces. She holds a master’s degree in computer science and a bachelor’s degree in physics, and has additional academic experience in Philosophy and ML. She is passionate about emerging technology and cofounded Manchester Futurists, a thriving intellectual community group aiming to explore the social impact of technology and shape a positive future.
Michael Evans
bio
Michael Evans is a user experience research lead in Future Experience Technologies at BBC Research and Development, working with the other human-computer interaction (HCI) researchers and engineers to invent the public service media of the future. He has a lot of experience in developing tools and research methods for professional creative tasks, including directing 360° video, intelligible machine learning, and quality of experience evaluation. Before joining the BBC in 1999, he was a lecturer at the University of Reading, Reading, U.K., co-founding the Signal Processing Lab and leading HCI research. He is a chartered engineer and has completed a DPhil in spatial audio and psychoacoustics with BT Labs, Martlesham, U.K., in 1997.
Ronan Forman
bio
Ronan Forman is a software engineer working at the BBC. After studying computer science and artificial intelligence at the University of York, York, U.K., he joined the BBC Software Engineering Graduate Scheme in 2017, where he worked across the BBC, including in Research and Development. While in Research and Development, he worked on prototyping the Ed system for automatically editing videos based on artificial intelligence (AI) learned rules, and running a framing study to evaluate the quality of automated and professional output.
James Gibson
bio
James Gibson is a research and development engineer at the BBC Research and Development, U.K., who has worked on projects looking at how to transition broadcasters from serial digital interface to Internet Protocol (IP) the corresponding security requirements and new production opportunities, as well as new immersive applications that are enabled by 5G such as augmented reality/virtual reality (AR/VR) remote rendering.
Stephen Jolly
bio
Stephen Jolly leads the artificial intelligence (AI) in Media Production project at BBC Research and Development, which is exploring how AI and machine learning will transform the media industry. He is a strong advocate of the potential for Intelligent Cinematography to help automate and democratise the production of film and television, and is always interested to hear from potential partners with an interest in the field. In his career at the BBC, he has also worked on a wide range of other technologies—from 3D and high frame-rate television to multidevice media and the Internet of Things. He joined the BBC in 2004, following the successful completion of a PhD in high energy physics at Imperial College, London and CERN, Geneva. He also holds a BSc in physics from Imperial College, London, U.K.
Lianne Kerlin
bio
Lianne Kerlin is a research scientist at BBC Research and Development who is interested in the impact of technology on people and society. She leads research and development work around human values; a project that translates psychological insight into actionable tools to shape future media services with people’s values at the heart of the innovation process. She is currently focused on turning human values into psychometrics, and by doing so provide new ways to measure the impact and value of digital services beyond the standard consumption metrics.
Susan Lechelt
bio
Susan Lechelt is a post-doctoral researcher in creative informatics at The University of Edinburgh, Edinburgh, U.K. Her research involves engaging with creative practitioners in art, film, design, and beyond, to help them envision how new forms of technology and data-driven innovation might fit into their future practice. Before joining the University of Edinburgh, she completed a PhD in human-computer interaction at University College London (UCL), in collaboration with BBC Research and Development.
Graeme Phillipson
bio
Graeme Phillipson currently works in BBC Research and Development on artificial intelligence (AI) in production project. This project is investigating how BBC can use techniques from AI, machine learning, and computer vision to tackle some of the problems found in television production. Previously in the BBC, Phillipson has worked on several projects involved in the running of BBC iPlayer, a system which automatically keeps track of what music has been played on TV and radio so that audiences can find music they have heard. Prior to working at the BBC, he studied neuroscience at Edinburgh University, and worked on a 3D motion capture system used in clinical gait analysis.
Matthew Shotton
bio
Matthew Shotton is the founder and chief executive officer of Algoloop, a human in the loop machine learning (ML) consultancy based in Berkeley, CA. He previously worked as a research engineer at BBC Research and Development, where he worked on ML systems for automatically editing media content, interactive video rendering engines, and a range of other projects. He is passionate about the intersection of art and technology and is an avid tinkerer and hobbyist with a degree in electronic engineering from Manchester University, Manchester, U.K.
Keyword(s)
Broadcast, Intelligent Cinematography, User Evaluation
Copyright
© 2020 SMPTE

Bibliographic Reference(s)

Source Data (JSON)

Full registry record with provenance metadata. Open directly: /api/doc/10.5594-JMI.2020.2967204.json

Reference Tree

Explore all references and references to this document, as a navigable tree.

Open Reference Tree
Reference this Doc

Plain text (ISO 690 compliant)

Preview:
Craig Wright, Jack Allnutt, Rosie Campbell, Michael Evans, Ronan Forman, James Gibson, Stephen Jolly, Lianne Kerlin, Susan Lechelt, Graeme Phillipson, and Matthew Shotton; AI IN PRODUCTION: VIDEO ANALYSIS AND MACHINE LEARNING FOR EXPANDED LIVE EVENTS COVERAGE, SMPTE Motion Imaging Journal ( Volume: 129, Issue: 2, March 2020); SMPTE, 2020. Available at https://doi.org/10.5594/JMI.2020.2967204
Snippet:
Craig Wright, Jack Allnutt, Rosie Campbell, Michael Evans, Ronan Forman, James Gibson, Stephen Jolly, Lianne Kerlin, Susan Lechelt, Graeme Phillipson, and Matthew Shotton; AI IN PRODUCTION: VIDEO ANALYSIS AND MACHINE LEARNING FOR EXPANDED LIVE EVENTS COVERAGE, SMPTE Motion Imaging Journal ( Volume: 129, Issue: 2, March 2020); SMPTE, 2020. Available at https://doi.org/10.5594/JMI.2020.2967204

HTML (ISO 690 compliant)

Preview:
Craig Wright, Jack Allnutt, Rosie Campbell, Michael Evans, Ronan Forman, James Gibson, Stephen Jolly, Lianne Kerlin, Susan Lechelt, Graeme Phillipson, and Matthew Shotton; AI IN PRODUCTION: VIDEO ANALYSIS AND MACHINE LEARNING FOR EXPANDED LIVE EVENTS COVERAGE, SMPTE Motion Imaging Journal ( Volume: 129, Issue: 2, March 2020); SMPTE, 2020. Available at https://doi.org/10.5594/JMI.2020.2967204
Snippet:
<span class="citation">Craig Wright, Jack Allnutt, Rosie Campbell, Michael Evans, Ronan Forman, James Gibson, Stephen Jolly, Lianne Kerlin, Susan Lechelt, Graeme Phillipson, and Matthew Shotton; <cite>AI IN PRODUCTION: VIDEO ANALYSIS AND MACHINE LEARNING FOR EXPANDED LIVE EVENTS COVERAGE</cite>, SMPTE Motion Imaging Journal ( Volume: 129, Issue: 2, March 2020); SMPTE, 2020. Available at <a href="https://doi.org/10.5594/JMI.2020.2967204" target="_blank" rel="noopener">https://doi.org/10.5594/JMI.2020.2967204</a></span>

SMPTE Icon SMPTE's HTML Pub

Preview:
Craig Wright, Jack Allnutt, Rosie Campbell, Michael Evans, Ronan Forman, James Gibson, Stephen Jolly, Lianne Kerlin, Susan Lechelt, Graeme Phillipson, and Matthew Shotton; AI IN PRODUCTION: VIDEO ANALYSIS AND MACHINE LEARNING FOR EXPANDED LIVE EVENTS COVERAGE, SMPTE Motion Imaging Journal ( Volume: 129, Issue: 2, March 2020); SMPTE, 2020
doi: 10.5594/JMI.2020.2967204
url: https://doi.org/10.5594/JMI.2020.2967204
Snippet:
<li>
Craig Wright, Jack Allnutt, Rosie Campbell, Michael Evans, Ronan Forman, James Gibson, Stephen Jolly, Lianne Kerlin, Susan Lechelt, Graeme Phillipson, and Matthew Shotton; <cite id="bib-10-5594-jmi-2020-2967204">AI IN PRODUCTION: VIDEO ANALYSIS AND MACHINE LEARNING FOR EXPANDED LIVE EVENTS COVERAGE</cite>, SMPTE Motion Imaging Journal ( Volume: 129, Issue: 2, March 2020); SMPTE, 2020
<span class="doi">10.5594/JMI.2020.2967204</span>
</li>