NAB 2019: TVU Networks’ MediaMind Merges Time Code, AI-Based Facial/Speech Recognition for Video Retrieval
TVU Networks (Booth C1707) will announce the new MediaMind Appliance, a metadata generation and search engine that speeds production performance and content reuse within an existing broadcast media supply chain. The system uses a combination of time-code and AI-based facial and speech recognition to generate metadata. And an open API lets users connect directly with MAM and NLE systems.
“TV stations and news agencies have considerable media assets, but how can they be considered assets if they can’t be reused and repurposed?” says Paul Shen, CEO, TVU Networks. “While there are many ways to measure the value of content in the media industry, there are ultimately only two: who gets to the audience first and what content is viewed and reused the most? The faster the content spreads, the more it is reused and repurposed, the greater the value of the content and the lower the production cost. That’s what’s driven the creation of the TVU MediaMind appliance and the entire MediaMind workflow.”
Because the MediaMind appliance utilizes both AI and time code to generate search results, it can convert audio and video into structured data in real-time. It’s capable of searching a raw asset at any stage of production – from assignment through to on-air. With the help of TVU’s AI engine, real-time speech recognition lets users find text results even faster than they would searching for images alone. Users simply type in what they’re looking for and the appropriate section of a time code appears. The real-time content search engine uses time code to identify each asset with a unique ID.
Clips can be marked and saved for immediate use. There’s no need to wait for pre-production to finish or for the video file to be entered into the MAM system. With the reporter still on scene, the clip is ready to go online or on-air.
The TVU MediaMind appliance is part of a TVU SaaS model that uses TVU MediaMind server (dual-channel VS3500/VS3100 models) as the edge server. The servers can receive video signals from any TVU Networks’ transmission device as well as process IP streams of standard transmission protocol, SDI video, and store files in any MAM system. They are professional video hardware decoders designed for the centralized viewing, control, and distribution of live video content.
In addition to the MediaMind appliance, the MediaMind ecosystem includes the following cloud-based products: TVU Producer for multi-camera live video production; TVU Remote Production System (RPS) for remote production using a broadcaster’s existing studio control room and a public Internet connection from the field; TVU CAS (Contribution Automation Solution) for the identification of all video acquisition resources on a given story at acquisition; TVU Grid – an IP-based video switching, routing and distribution solution; and TVU Grid Market, a central source for content owners to contribute and pull video enabling assets to be shared and sold on a global scale.
TVU MediaMind Appliance Highlights:
- Generates metadata for each video segment based on AI for facial and speech recognition and video time code
- Realize assets value with immediate reuse and repurposing across platforms, programs and organizations
- Exposure of assets to global media market
- Saves expensive hardware cost
- Reduces operation and maintenance cost
- Open API lets users connect with MAM and NLE systems
- No laborious cataloging required
- Includes the latest integration and support for automation, AI, streaming, 4K, and 5G