Artificial intelligence AI - Machine learning ML
Artificial Intelligence (AI) plays an emerging but significant role in the media and entertainment world. It can be leveraged to automate functions in workflows, analyse data for targeted advertising and content recommendations, and even amplify streaming quality by pre-positioning video assets closer to the subscribers. It means that users can stream high-quality video even during peak hours.
AI can also be harnessed to influence the all-important compression elements of a content workflow. Compression is a fundamental part of any content workflow. The ultimate goal of compression? To achieve the highest possible video quality, at the lowest possible bitrates, and all locked within a defined compute budget. Compression, however, can be very compute-intensive, requiring a lot of processing power and bitrate efficiency. In terms of enabling more sustainable content creation, compression plays an increasing role in reducing power and ensuring best density to reduce server footprint and power consumption.
This is why MediaKind’s dedicated in-house compression research team has developed a pioneering technology that can help make a real difference in this area. MediaKind’s new intuitive compression algorithm, AI-based Technology Compression (ACT), enables substantial operational and performance improvements based on solid science-based research. AI-based technology allows us to analyse the type of content that arrives at the encoder in real-time and drive the processing resource allocation towards different tools. ACT is a crucial and innovative technology that can enhance the impact of compression cogs within a content-production machine. AI is used to select the most effective balance of compression methods based on the nature of the content and map how the encoder uses its processing resources accordingly. This is possible for both traditional linear and ABR encoding and transcoding.
With ACT, users benefit from optimal processing thanks to a better, dynamic, choice of processing for groups of tools, producing lower bitrates for a given amount of processing resource. The MediaKind algorithm also adapts to the available resource to improve density. For example, when running a standard encode mode for four HD channels, allocating all of the available CPU resources evenly per channel, ACT could allow users to run six channels instead – all within the same infrastructure. In doing so, users could increase the channel density by up to 50% within a situation where encode density is the most important consideration.
ACT also provides greater flexibility as users can spin up new channels without the need for new infrastructure or reconfiguring existing services. Operational setup is also simplified, and infrastructure sizing is made much easier, meaning there is no need to select profiles ahead of time. ACT can automatically expand the scope of compression tools to always take advantage of all processing power, tailored to the content as it changes. For example, two conventional encoder profiles on a particular server or node size might leave a significant amount of potential processing power untapped. Using ACT will extend the utilisation for the best performance.
Leveraging the elasticity enabled by cloud-based deployments means users can set a specific but configurable CPU limit for application deployment, making cloud resource management easier.
ACT is already integrated within several of MediaKind’s solutions, powering offerings such as Aquila Broadcast, Aquila Streaming and MK Encoding Live. The technology has proven impacts on enabling cutting-edge media experiences. In all three instances, ACT helps to deliver superior compression efficiency. The video processing is used optimally without exceeding allocations producing lower bitrates for a given amount of processing resource. ACT can also improve the density of a server due to the algorithm adapting to fit within the available resource. The technology also adds operational value for these MediaKind solutions and products as the need to select profiles ahead of time are removed. Instead, users can set specific and configurable application deployments in the cloud. These advantages lead to a significant operational cost-benefit for the end-user determined by the scale and type of their media content processing deployment.