Internet Delivered Media - Trends & Best Practices
As part of this tutorial, we will give an overview as well as best practices for playback and creation of adaptive bitrate (ABR) content. With streaming formats such as Dynamic Adaptive Streaming over HTTP (MPEG-DASH) and HTTP Live Streaming (HLS), content providers can reach many devices (mobile, desktop, TV, etc.) over-the-top (OTT). The MPEG Common Media Application Format (CMAF) standard will enable interoperability between both streaming formats by leveraging the same media format.HTML5 APIs Media Source Extensions (MSE) and Encrypted Media Extensions (EME) enable playback interoperability across all browser-based platforms.Latest advancements in video encoding allow for title-based encoding, making traditional, generic encoding ladders obsolete and thereby saving CDN costs while boosting the quality of experience.
In order to distribute premium content, Digital Rights Management (DRM) and watermarking solutions are needed to protect the media streams. A multi-DRM approach is required in order to protect content with more than one DRM system - the MPEG Common Encryption (CENC) standard enables this. Many entities are involved in a multi-DRM backend, e.g. license server, encryptor or packager. The Content Protection Information Exchange Format (CPIX) specification enables secure and interoperable communication between the entities.
In addition, advanced media streaming features such as streaming analytics, client coordination, ad-insertion, two-step watermarking, low latency streaming and 360° streaming are necessary to provide an optimal and enhanced streaming experience.
This tutorial will cover the aforementioned latest enhancements in internet-delivered media. We will demonstrate the key technologies in multiple live demos and identify challenges and potential hazards when applying the technologies in production.
Foundations of Adaptive Streaming
- Streaming Formats: DASH, CENC, HLS, CMAF
- Web Media APIs: MSE/EME, fetch(), XHR, WebSocket
- ABR Algorithms
- Standards update: DASH-IF, WAVE etc.
- Encoding & Player Best Practices
- Cross-platform deployment to SmartTVs, HbbTV, FireTV, Chromecast, AppleTV, iOS, Android, Desktop etc.
- Feature support: Codecs, Casting, Subtitles, Parental control etc.
- Multi-DRM backend with CPIX
- Web, native or hybrid apps?
- Deployment Best Practices
Advanced Web Media Streaming
- Streaming Metrics Analytics using SAND
- OTT and HbbTV Ad-insertion
- Two-step Watermarking
- Low-latency streaming
- 360° Streaming
Lecturer: Stefan Pham
HbbTV - deep dive and hands-on
This tutorial gives an overview of the Hybrid Broadcast Broadband Television Standard (HbbTV) with particular focus on the latest HbbTV 2 specification.
In the first part of the tutorial we will give a short introduction about the HbbTV standard highlighted with some practical examples and real-world applications. In the second part of the tutorial we will present the most relevant features introduced in HbbTV 2 such as the new HTML5 APIs, companion screens and media synchronisation. In contrast to the new HbbTV 2 features, we will discuss new use cases and applications that are enabled by these features. Afterwards, we will present HbbTV-related specifications like Operator Applications and Application Discovery over Broadband. The last part of the tutorial gives a overview on tools and frameworks that can be used for developing HbbTV application.
Introduction in HbbTV
- What is HbbTV
- How it Works
- History and Evolution of the Versions
- Real World HbbTV Apps
- HTML5 associated technologies: CSS3, DOM3, HTML5 Video/Audio, Canvas 2D, Web Sockets, Web Storage, Web Audio
- Subtitles: EBU profile of W3C TTML
- Companion Screen: Discovery, Launch and App2App Communication
- Multi-stream Synchronization, Multi-device Synchronization and Application & content synchronisation
- Updated DVB-DASH: High Dynamic Range Video (HDR), High Frame Rate video (HFR), Next Generation Audio (NGA)
- HEVC Video
- Non-realtime Content Delivery via Broadcast
- Multiple HTML5 Video Elements
- Support of Additional Input Devices (Mouse, Keyboard)
HbbTV 2 Enabled Use Cases
- Advert Insertion into VoD content
- Advert Overlays Synchronized with Broadcast
- Push VoD
- Alternative or Personalized Audio on Companion Screen
- Personal Programme Guide and Additional Information on Companion Screen
- Play-along experience (Quiz show) and Multi-player Games
- Push to Mobile
- 360° Video
Other HbbTV related Specs
- Operator Applications (OpApps)
- Application Discovery over Broadband
- IP-Delivered Broadcast Channels and Related Signalling of HbbTV Applications
- Which Dev Tools to use
- HbbTV Libraries and Frameworks
- How to Debug and Test HbbTV Apps
- HbbTV Tools for non-programmers (MPAT)
Lecturer: Louay Bassbouss
Recommender Systems and Machine Learning
A set of different concepts and terms arose in the last few years: Big Data, Artificial Intelligence, Recommender Systems, Data Mining, Data Science, Machine Learning, Deep Learning and Neuronal Networks – just to mention a few of them. What is the difference and what do they have in common? This tutorial gives an overview of the major techniques and focusses on one particular method for personalization: Recommender Systems.
Currently, most users are not able to consume all offered items at once – nor in a life time. So they have to find a considered selection of those items they want to buy or consume. A recommender system helps its users to decide for products or media they might be interested in. Therefore, a set of different approaches can be utilized to get predictions of the users' behaviors and preferences.
We will present different application areas for artificial intelligence in the modern connected world – with a special focus on these applications that make the users’ decisions more convenient, efficient and effective. Some basic concepts will be introduced and explained with simple and comprehensible examples of algorithms. Moreover, the tutorial addresses some common challenges as well as typical issues and discusses how service providers might overcome them.
The tutorial is of interest for you if…
- … you want to learn why some recommender systems present the same product again, and again and again – even after you already purchased it!
- … you are wondering why it is sometimes better to hire 40 employees than to apply 40 additional algorithms to improve your predictions.
- … you want to get a general understanding of machine learning and recommender systems or if you want to develop some ideas how to personalize your web service.
Lecturer: Christopher Krauß