We know AI is going to stick around. Whether it’s AI, Machine Learning, Deep Learning or by another name, it all stacks up to the same thing: we’re breaking away from fixed algorithms where one equation ‘does it all’ to a much more nuanced approached with a better result. This is true across all industries. Within the Broadcast industry, one way it can be used is in video and audio compression. Want to make an image smaller? Downsample it with a Convolutional Neural Network and it will look better than Lanczos. No surprise, then, that this is coming in full force to a compression technology near you.
In this talk from Comcast’s Dan Grois, we hear the ongoing work to super-charge the recently released VVC by replacing functional blocks with neural-networks-based technologies. VVC has already achieved 40-50% improvements over HEVC. From the work Dan’s involved with, we hear that more gains are looking promising by using neural networks.
Dan explains that deep neural networks recognise images in layers. The brain does the same thing having one area sensitive to lines and edges, another to objects, another part of the brain to faces etc. A Deep Neural Network works in a similar way.
During the development of VVC, Dan explains, neural network techniques were considered but deemed too memory- or computationally-intensive. Now, 6 years on from the inception of VVC, these techniques are now practical and are likely to result in a VVC version 2 with further compression improvements.
Dan enumerates the tests so far swapping out each of the functional blocks in turn: intra- and inter-frame prediction, up- and down-scaling, in-loop filtering etc. He even shows what it would look like in the encoder. Some blocks show improvements of less than 5%, but added together, there are significant gains to be had and whilst this update to VVC is still in the early stages, it seems clear that it will provide real benefits for those that can implement these improvements which, Dan highlights at the end, are likely to require more memory and computation than the current version VVC. For some, this will be well worth the savings.
“Enhance!” the captain shouts as the blurry image on the main screen becomes sharp and crisp again. This was sci-fi – and this still is sci-fi – but super-resolution techniques are showing that it’s really not that far-fetched. Able to increase the sharpness of video, machine learning can enable upscaling from HD to UHD as well as increasing the frame rate.
Adithyan outlines two main enablers of super-resolution, allowing it to displace the traditional methods such as bicubic and Lanczos. Enabler one is the advent of machine learning which now has a good foundation of libraries and documentation for coders allowing it to be fairly accessible to a wide audience. Furthermore, the proliferation of GPUs and, particularly for mobile devices, neural engines is a big help. Using the GPUs inside CPUs or in desktop PCI slots allows the analysis to be done locally without transferring great amounts of video to the cloud solely for the purpose of processing or identification. Furthermore, if your workflow is in the cloud, it’s now easy to rent GPUS and FPGAs to handle such workloads.
Using machine learning doesn’t only allow for better upscaling on a frame-by-frame basis, but we are also able to allow it to form a view of the whole file, or at least the whole scene. With a better understanding of the type of video it’s analysing (cartoon, sports, computer screen etc.) it can tune the upscaling algorithm to deal with this optimally.
Anime has seen a lot of tuning for super-resolution. Due to Anime’s long history, there are a lot of old cartoons which are both noisy and low resolution which are still enjoyed now but would benefit from more resolution to match the screens we now routinely used.
Adithyan finishes by asking how we should best take advantage of super-resolution. Codecs such as LCEVC use it directly within the codec itself, but for systems that have pre and post-processing before the encoder, Adithyan suggests it’s viable to consider reducing the bitrate to reduce the CDN costs knowing the using super-resolution on the decoder, the video quality can actually be maintained.
Today’s video has a wide array of salient topics from seven speakers at SMPTE Toronto’s meeting in February. Covering Uncompressed IP networking, colour theory & practice, real-time virtual studios and AI, those of us outside of Toronto can be thankful it was recorded.
Ryan Morris from Arista (starting 22m 20s) is the first guest speaker and kicks off with though-provoker: showing the uncompressed bandwidths of video, we see that even 8K video at 43Gb/s is much lower than the high-end network bandwidths available in 400Gbps switch ports available today with 800Gbps arriving within a couple of years. That being said, he gives us an introduction to two of the fundamental technologies enabling the uncompressed IP video production: Multicast and Software-Defined Networking (SDN).
Multicast, Ryan explains is the system of efficiently distributing data from one source to many receivers. It allows a sender to only send out one stream even if there are a thousand receivers on the network; the network will split the feed at the nearest common point to the decoder. This is all worked out using the Internet Group Message Protocol (IGMP) which is commonly found in two versions, 2 and 3. IGMP enables routers to find out which devices are interested in which senders and allows devices to register their interest. This is all expressed by the notion of joining or leaving a multicast group. Each multicast group is assigned an IP address reserved by international agreement for this purpose, for instance, 184.108.40.206 is one such address.
Ryan then explores some of the pros and cons of IGMP. Like most network protocols each element of the network makes its own decision based on standardised rules. Though this works well for autonomy, it means that there no knowledge of the whole system. It can’t take notice of link capacity, it doesn’t know the source bandwidth, you can guess where media will flow, but it’s not deterministic. Broadcasters need more assurance of traffic flows for proper capacity planning, planned maintenance and post-incident root cause analysis.
Reasons to consider SDN over IGMP
SDN is an answer to this problem. Replacing much of IGMP, SDN takes this micro-decision making away from the switch architecture and replaces it with decisions made looking at the whole picture. It also brings an in important abstraction layer back to broadcast networks; engineers are used to seeing X-Y panels and, in an emergency, it’s this simplicity which gets things back on air quickly and effectively. With SDN doing the thinking, it’s a lot more practical to program a panel with human names like ‘Camera 1’ and allow a take button to connect it to a destination.
Next is Peter Armstrong from THP who talks about colour in television (starting 40m 40s). Starting back with NTSC, Peter shows the different colour spaces available from analogue through to SD then HD with Rec 709 and now to 3 newer spaces. For archiving, there is an XYZ colour space for archival which can represent any colour humans can see. For digital cinema there is DCI-P3 and with UHD comes BT 2020. These latter colour spaces provide for display of many more colours adding to the idea of ‘better pixels’ – improving images through improving the pixels rather than just adding more.
Another ‘better pixels’ idea is HDR. Whilst BT 2020 is about Wide Colour Gamut (WCG), HDR increases the dynamic range so that the brightness of each pixel can represent a brightness between 0 and 1000 NITs, say instead of the current standard of 0 to 100. Peter outlines the HLG and PQ standards for delivering HDR. If you’re interested in a deeper dive, check out our library of articles and videos such as this talk from Amazon Prime Video. or this from SARNOFF’s Norm Hurst.
ScreenAlign device from DSC Labs
SMPTE fellow and founder of DSC Laboratories, David Corley (56m 50s), continues the colour theme taking us on an enjoyable history of colour charting over the past 60 years up to the modern day. David explains how he created a colour chart in the beginning when labs were struggling to get colours correct for their non-black and white film stock. We see how that has developed over the years being standardised in SMPTE. Recently, he explains, they have a new test card for digital workflows where the camera shoots a special test card which you also have in a digital format. In your editing suite, if you overlay that file on the video, you can colour correct the video to match. Furthermore, DSC have developed a physical overlay for your monitor which self-illuminates meaning when you put it in front of your monitor, you can adjust the colour of the display to match what you see on the chart in front.
Gloria Lee (78m 8s) works for Graymeta, a company whose products are based on AI and machine learning. She sets the scene explaining how broadly our lives are already supported by AI but in broadcast highlights the benefits as automating repetitive tasks, increasing monetisation possibilities, allowing real-time facial recognition and creating additional marketing opportunities. Gloria concludes giving examples of each.
Cliff Lavalée talks about ‘content creation with gaming tools’ (91m 10s) explaining the virtual studio they have created at Groupe Média TFO. He explains the cameras the tracking and telemetry (zoom etc.) needed to ensure that 3 cameras can be moved around in real-time allowing the graphics to follow with the correct perspective shifts. Cliff talks about the pros and cons of the space. With hardware limiting the software capabilities and the need for everything to stick to 60fps, he finds that the benefits which include cost, design freedom and real-time rendering create an over-all positive. This section finishes with a talk from one of the 3D interactive set designers who talks us through the work he’s done in the studio.
Mary Ellen Carlyle concludes the evening talking about remote production and esports. She sets the scene pointing to a ‘shifting landscape’ with people moving away from linear TV to online streaming. Mary discusses the streaming market as a whole talking about Disney+ and other competitors currently jostling for position. Re-prising Gloria’s position on AI, Mary next looks further into the future for AI floating the idea of AI directing of football matches, creating highlights packages, generating stats about the game, spotting ad insertion opportunities and more.
Famously, Netlflix has said that Fortnite is one of its main competitors. And indeed, esports is a major industry unto itself so whether watching or playing games, there is plenty of opportunity to displace Netflix. Deloitte Insights claim 40% of gamers watch esports events at least once a week and in terms of media rights, these are already in the 10s and 100s of millions and are likely to continue to grow. Mary concludes by looking at the sports rights changing hands over the next few years. The thrust being that there are several high profile rights auctions coming up and there is likely to be fervent competition which will increase prices. Some are likely to be taken, at least in part, by tech giants. We have already seen Amazon acquiring rights to some major sports rights.
Artificial Intelligence and Machine Learning (ML) dominate many discussions and for good reason, they usually reduce time and reduce costs. In the broadcast industry their are some obvious areas where it will, an already does, help. But what’s the time table? Where are we now? And what are we trying to achieve with the technology?
Edmundo Hoyle from TV Globo explains how they have managed to transform the thumbnail selection for their OTT service from a manual process taking an editor 15 minutes per video to an automated process using machine learning. A good thumbnail is relevant, it is a clear picture and has no nudity or weapons in it. Edmundo explains that they tackled this in a three-step process. The first step uses NLP analysis of the episode summary to understand what’s relevant and to match that with the subtitles (closed captions). Doing this identifies times int he video which should be examined more closely for thumbnails.
The durations identified by the process are then analysed for blur-free frames (amongst other metrics to detect clear videography) which gives them candidate pictures which may contain problematic imagery. the AWS service Rekognition which returns information regarding whether faces, guns or nudity are present in the frame. Edmundo finishes by showing the results which are, in general very positive. Final choice of thumbnails is still moderated by editors, but the process is much more streamlined because they are much less likely to have to find an image manually since the process selects 4 options. Edmundo finishes by explaining some of the chief causes of rejecting an image which are all relatively easy to improve upon and tend to be related to a person looking down or away from the camera.
We’ve seen before on The Broadcast Knowledge the idea of super-resolution which involves up-scaling images/video using machine learning. The result is better than using standard linear filters like lanczos. This is has been covered in a talk from Mux’s Nick Chadwick about LCEVC. Yiannis Andreopoulos from iSize talks next about the machine learning they use to improve video which uses some of these same principles to pre-treat, or as they call it ‘pre-code’ video before it’s encoded using a standard MPEG encoder (whether that be AVC, HEVC or the upcoming VVC). Yiannis explains how they are able to understand the best resolutions to encode at and scale the image intelligently appropriately. This delivers significant gains across all the metrics leading to bandwidth reduction. Furthermore he outlines a system which feeds back to maintain both the structure of the video which avoids it becoming too blurry which can be a consequence of being to subservient to the drive to reduce bitrate and thus simplifying the picture. It can also, though, protect itself from going too far down the sharpness path and only chasing metrics gains. He concludes by outlining future plans.
Grant Franklin Totten then steps up to explain how Al Jazeera have used AI/machine learning to help automate editorial compliance processes. He introduces the idea of ‘Contextual Video Metadata’ which ads a level of context to what would otherwise be stand-alone metadata. To understand this, we need to learn more about what Al Jazeera is trying to achieve.
As a news organisation, Al Jazeera has many aspects of reporting to balance. They are particularly on the look out for bias, good fact-checking & fake news. In order to support this, they are using AI and machine learning. They have both textual and video-based methods of detecting fake news. As an example of their search for bias, they have implemented voice detection and analysed MP’s speech time in Ireland. Irish law requires equal speaking time, yet Al Jazeera can easily show that some MPs get far more time than others. Another challenge is detecting incorrect on-screen text with the example given of naming Trump as Obama by accident on a lower-third graphic. Using OCR, NLP and Face recognition, they can flag issues with the hope the they can be corrected before Tx. In terms of understanding, for example, who is president, Al Jazeera is in the process of refining the Knowledge graph to capture the information they need to check against.
AI and machine learning (ML) aren’t going anywhere. This talk shines a light on two areas where it’s particularly helpful in broadcast. You can count on hearing significant improvements in AI and ML’s effectiveness in the next few years and it’s march into other parts of the workflow. Watch now! Speakers
Grant Franklin Totten
Head of Media & Emerging Platforms,
Al Jazeera Media Networks
Edmundo Hoyle (GLOBO), Yiannis Andreopoulos (iSize Technologies) and Grant Totten (Al Jazeera Media Network).
Subscribe to get daily updates
Views and opinions expressed on this website are those of the author(s) and do not necessarily reflect those of SMPTE or SMPTE Members.
This website is presented for informational purposes only. Any reference to specific companies, products or services does not represent promotion, recommendation, or endorsement by SMPTE