What lies beneath: How data transfer enables AI

0
80
Written by Richard Heitmann, VP & General Manager, Aspera

SCREEN AFRICA EXCLUSIVE:

Artificial Intelligence (AI) is impacting business, technology and everyday life at a dizzying pace. Data is the fuel of AI.

An enormous amount of data must be captured, stored and analysed before it’s turned into actionable insights that inform decision-making and drive a wide range of technical applications. At the same time, collection, analysis and storage capabilities have improved dramatically in recent years. So has the means to transport data, an essential AI enabler in the media and entertainment space.

Files and data must be transferred and manipulated to make AI applications possible. Yet standard TCP (transmission control protocol)-based transport applications, traditionally used to transfer data over IP networks, lack the speed and throughput reliability to securely transfer large volumes of data in practical time-frames. Advances in transfer technology have opened the door for new and innovative media applications. A best-in-class high-speed data transfer service can enable near-real time AI workflows by delivering content and associated metadata across global distances at maximum speed. This makes it possible to access structured information from video files, speech transcriptions and speech-synthesis for MAM system categorisation, clips creation, content personalisation and more.

Transport protocol for a new age

The inferior performance of legacy transfer applications such as FTP (file transfer protocol) and HTTP (hypertext transfer protocol) are a direct result of TCP’s inherent limitations. A better approach exists; one that solves big data and media transport issues by eliminating TCP’s underlying bottleneck altogether. Aspera FASP is a bulk data transport technology implemented at the application layer, that provides secure high-speed transfer while remaining compatible with existing networks and infrastructure.

Leveraging existing WAN (wide area network) infrastructure and commodity hardware, the technology fully utilises available network bandwidth, transferring data at high speeds regardless of network conditions and transfer distance. The protocol can achieve speeds up to hundreds of times faster than FTP and HTTP, with end-to-end security, guaranteed reliability and bandwidth control. It can also scale up as needed, limited only by the resources of the end-point computers. Bandwidth management provides transfer prioritisation and automatic protection of other IP traffic. Predictable throughput is independent of network delay and resilient to extreme packet loss regardless of network conditions, delivering content and files across distance at maximum speed and enabling near-real time AI workflows.

Meta magic: unlocking metadata

This high-speed data transfer enables a range of AI engines, like IBM Watson Media — an artificial intelligence platform that understands, reasons and learns. Integration with Aspera Orchestrator, a web-based application and SDK for workflow automation, enables the ultra-high-speed ingest and high-volume processing needed for AI workflows. These capabilities, along with base technologies built on machine learning principles, drive video enhancement, allowing Watson to engage without additional investment in satellite uplinks, dedicated fibre, onsite equipment or manpower. AI-driven capabilities like speech-to-text, natural language understanding, sentiment and tone analysis and semantic scene detection create searchable metadata, bringing powerful new capabilities to video. Users — from broadcasters to content owners and rights holders — can automate tasks and deliver more customised content and engaging viewer experiences.

With a deeper understanding of their content, media organisations are improving search capabilities, pulling more specific clips from their content libraries and making more granular and targeted content recommendations. The ability to deliver the right content to the right viewer at the right time is an increasingly invaluable marketplace skill.

Automation and beyond

In addition to greater search, automation and targeting capabilities, there’s sure to be more integration across media experiences and broader use of digital assistance to aid content discovery. All of this begs the question, When is a human needed to intervene, to check and monitor and to craft content?

The biggest value of AI is its ability to help scale human effort. As a supplementary technology, AI solutions empower content creators and editors to work more efficiently and make more informed business decisions. But, at the end of the day, humans have the ultimate power and sign-off control.

AI has already significantly changed the media landscape — both in the manner in which audiences discover and engage with content and in the way that content is created and distributed. The combination of high-speed data transport and AI is sure to further impact creative and production workflows in exciting new ways. Currently, algorithms not only influence what audiences see on different platforms, but also the content that is being created, fundamentally changing audience behaviour and the creative process. AI and high-speed data transport may well be rewriting the rules for creating and acquiring content.

Written by Richard Heitmann, VP & General Manager, Aspera – an IBM company.

LEAVE A REPLY

Please enter your comment!
Please enter your name here