Unlocking the Secrets of Digital Media Vaults with AI

May 1, 2024 · 5 min read
Blurred Shape
Blurred Shape
Blurred Shape
Blurred Shape

By Jason Perr

Unlocking the Secrets of Digital Media Vaults with AI

On October 14, 1888, in Northern England, French inventor Louis Le Prince created what would become known as the oldest surviving motion picture with a runtime of just over one and a half seconds. 136 years later we are seeing over 500 hours of new video uploaded to YouTube alone every minute of every day.

In the 1920s, the motion picture studios became established and started to build their own movie libraries. While the preservation of film is complex with different requirements for various types of materials, it generally took the form of closets, unused offices, warehouses, and salt mines in Kansas, where film was stored on reels and stacked up or loosely organized. Many items from these historic collections have been lost due to degradation, whether it be from age, neglect, or just being lost track of at some point. Over the years as movies became a cultural phenomenon the world over, it became more and more difficult for the respective owners of these treasured historical artifacts to organize, track, and ensure they could find what they needed when they needed it.

The Rise of Film Preservation

While as early as the 1930s, there was recognition of the need to focus on film preservation, it was another 50 years before the field received official status with the United Nations Educational, Scientific and Cultural Organization (UNESCO) publishing the Recommendation for the Safeguarding and Preservation of Moving Images that declared “All moving images of national production should be considered by Member States as an integral part of their ‘moving image heritage.’”

 The importance of preservation efforts cannot be denied and the discussion of film archives vs. digital archives is sure to rage on amongst film purists and technologists. However, there is no doubt that the digital revolution was a big jump forward in the struggle to store information long term, and that includes digital assets.

Fast Forward to the Digital Revolution

For those of us born prior to the digital revolution that took place around the turn of the millennium, it is easy to recount the many times we’ve heard that lost footage had been found and it seems relatively commonplace today, spanning everything from The Beatles studio performances to Disney cartoons.

Sanity Starts with S3 Storage

As the digital revolution got underway, storage technologists started to wrestle with the challenge of how to store massive amounts of unstructured data. This gave rise to S3-compatible object-based storage—a concept at the heart of Perifery’s nearline and archive solutions.

Designed to address the diverse needs of modern data management, Perifery’s innovative solutions were purpose built to provide unrivalled security and scalability for organizations working with large volumes of digital content. Deployed by leading Media and Entertainment companies worldwide, Perifery’s storage solutions ensure assets remain preserved and protected for the long term.

Combining both simplicity and power, our tools enable you to effortlessly access your content from anywhere, at any time. This approach gives you a complete understanding of your content, empowering you to harness the full potential of your large media libraries.

Regardless of your reasons for selecting a media asset storage solution, there is one incredibly compelling reason to use S3—it allows you to organize your data with metadata. You remember metadata, right? It is literally “data about the data.”

Metadata Meets Its Match with the Perifery Intelligent Content Engine

When it comes to searching for and validating digital files, metadata has played the hero for many years and is still a critical component for a plethora of use cases. However, when it comes to searching for a particular film clip—whether it is to document history, illustrate creative concepts, or re-use existing assets to either save money or generate new income—it now has a powerful and flexible partner: Artificial Intelligence (AI). More specifically, the new Perifery Intelligent Content Engine (ICE). ICE gives you AI content understanding without metadata—complete, insightful and simple.

All You Have to Do is Ask!

Two weeks ago, I was at the National Association of Broadcasting Show (NAB Show) in Las Vegas. There, technology vendors from around the world gathered to talk to each other and potential partners and customers about media, entertainment, broadcasting, and everything that goes with it—the challenges, the problems, and how to prepare for the future.

I spent the entire week talking non-stop about the Perifery Intelligent Content Engine (ICE). ICE is comprised of a group of powerful AI agents, each with its own specialty and set of skills. Some agents are great at analyzing video content, another one performs transcription and understanding of audio, one is great at long-term memory operations, and one has the very special job of interacting with users. This Agent is the only one with a name. Her name is AIME, your Artificial Intelligence Multimodal Emissary. AIME personifies the friendly, all-knowing librarian to the film library (aka, the nearline active archive). AIME knows about all of the content of an entire organization. Not just where it is, but what it is, why it was created, how to find it, and what is important about it for your organization. Finding content in a traditional MAM system or through a traditional large-scale filesystem requires training, process, and documentation. With AIME, all you need to do is ask.

Perifery ICE Beta Trial Starting Soon

We are currently taking requests from customers of Perifery AI+ (FX version) for beta testing. If you are interested in being one of the first to try this new product and provide input, please complete the short application below.

Existing Customer
This field is for validation purposes and should be left unchanged.

Related Articles

Subscribe to updates

This field is for validation purposes and should be left unchanged.
linkedin facebook pinterest youtube rss twitter instagram facebook-blank rss-blank linkedin-blank pinterest youtube twitter instagram