Broadcast Beat Magazine 2017 NAB Show NY/SMPTE Special Edition - Page 62

“Predicting the Future Wth Media Analytics Services,” Symposium Co-chair Thomas will describe how media intelligence based on big data analytics can increase efficiency in journal- istic workflows and assist operations in monetiz- ing content more efficiently. Key takeaways for attendees will include how to enrich metadata and guide production decisions; how to ensure the reliability of AI-generated data and nurture the learning machines; and how to visualize and present large amounts of data in a way that makes workflows more efficient. The morning program will wrap with “Machine Translation of Timed Text: My Machine Can Read Faster Than Me,” presented by Chris Reynolds, vice president, global partnerships, technology and innovation, Deluxe Entertainment Services Group; and Greg Taieb, senior director, Sfera at Deluxe Entertainment Services Group. Reynolds and Taieb will present examples and case studies that demonstrate the limits of current machine translation technology while also examining how ML may be implemented to continually improve the quality of translations continually. Additionally, they will explore how the current state of machine translation can be leveraged through cloud-based web applications to aid human translators and bring higher levels of quality and efficiency to video localiza- tion workflows. Following a net- working lunch, the Symposium will reconvene with a two-part tutorial focused on ML and presented by SMPTE Fellow and Symposium C o - c h a i r Munson. In part one, “Machine Learning Foundations (Theory and Practice),” Munson will offer a foundation for digital media engineers in the theoretical underpinnings of ML programs: what they are, an introduction to class- es of both supervised and unsupervised learning algorithms, and examples of these techniques in practice covered by leading over-the-top (OTT) 62 • Broadcast Beat Magazine • www.broadcastbeat.com and media content distribution technologists. In part two, “Using Machine Learning to Predict Optimal Sources for Content in a Heterogeneous Network,” Munson will outline a new ML algorithm and implementation that applies continuous mul- tivariable regression to help predict the best (most optimal) source of content in a distributed and heterogeneous storage network. In the next session, “Microsoft’s New Video Analytics Platform — A Scalable Turnkey Cloud Service for Understanding What’s in Your Media,” Martin Wahl, principal program manager for Microsoft’s Azure Media Services, will take the audience through the capabilities of the compa- ny’s robust new AI-based video indexing service and provide a view into its underlying ML models and cognitive services. Konstantin Wilms, principal solutions architect at Amazon Web Services, will follow up with “Content Intelligence on AWS — Integrating Artificial Intelligence and Machine Learning Technologies Into Cloud-based Media Workflows.” Here, Wilms will take a deep dive into AI and ML in the cloud, illustrating how these technologies and services can best be used to enhance both n ܁д)ѕЁݽɭ̸)QM5AQ)Mͥմݥ)ՑݥѠ)͍ͥ)ѥѱqQAݕ)$P!܁$) ) )=ȁ ̻ͥt)Áݥ̴)́'éɽܴ)ɽѡ)䰁ɥٕ)ѡݕȁ)ݹф)ѡ)Ѽɽ́)չхѡфQݥɕ͕Ё)ѥٕ́ѡ$̰ͥ)Ց܁Ѽѥ锁фѡ)́յ́qɕٕȁ̻t)QMͥմ́ٽɥєЁѡ)M5AQՅQ ɕᡥѥ