Vizrt releases AI-driven updates for its MAM Viz One 8.1, with new capabilities to help content creators log, find, create and repurpose media faster with more flexibility and accuracy.
Vizrt is releasing AI-driven updates to its media asset management (MAM) system, Viz One 8.1, focusing on speed, intelligence and accuracy.
Some broadcasters and content owners today are finding traditional workflows are no longer adequate when dealing with growing volumes of raw footage, having reached the limits of manual shot logging and archive research. Others feel pressured to produce more content, faster, across more platforms in order to remain competitive.
In response, Viz One 8.1 now integrates with aiconix's DeepVA sovereign AI platform, automating key stages of media management and helping teams to move from time-consuming manual processes to managing media quickly and accurately.
Automation through Visual Understanding
With the integration, Viz One 8.1 gains AI-assisted visual recognition allowing users to rapidly locate specific content. By adding visual understanding to archives and workflows, automatic identification and tagging of faces, objects and scenes within video material becomes possible,
Instead of manual logging, teams can review metadata logged by AI, train and use custom models for unknown entities, and perform optional quality assurance reviews. Reliability is significantly improved as well. Vizrt claims to have recorded up to a 50% increase in logging accuracy, 10 times the speed in logging workflows, and five times quicker search and discovery.
The ability to efficiently surface the best clips, scenes and themes expands options for reuse, compliance and, from there, monetization of the content."Content creators need AI that works for their specific needs, not a one-size-fits-all solution," said Esther Arroyo, CMO at aiconix. "DeepVA's ability to train custom models and generate structured, machine-readable metadata allows Viz One users the opportunity to optimise their workflows and deliver more relevant, localised content to their audiences."
Faster research and clip discovery can also lead to fewer bottlenecks and greater creative flexibility for broadcasters and content creators working across news, sports and entertainment on any platform.
DeepVA
Viz One 8.1 builds on the system's containerized architecture designed to scale smoothly across on-premises, hybrid and cloud environments. DeepVA's AI services are deployed securely within each Viz One environment, supporting effective AI workflows without risking sensitive data leaving the infrastructure.
DeepVA gives a level of choice and control to users, who can build, tune and use AI datasets and custom models via a single API across various workflows. Models can run independently of geopolitical differences – keeping data sovereignty – and without requiring teams to use their data for training.
By adding intelligence to workflow engines, DeepVA creates the foundation for automating repetitive media tasks with AI-driven analysis, especially when custom, localised metadata is required. Aiconix also has its own live metadata standard and format for enriched live metadata, AMT (Aiconix Metadata Transport), for fast, AI-powered subtitling. It supplies timecoded information when needed, transforming live streams into real-time, multi-lingual experiences with minimal delay,
Pragmatic Approach
Together, Vizrt and DeepVA have taken steps toward a pragmatic approach to AI in media – that is, applying automation where it adds actual value, without unnecessary complexity. Viz One 8.1 aims to deliver the kind of AI that makes a measurable difference, saving the hours needed for manual tagging, finding the most relevant assets instantly, and helping content creators tell their stories, faster.
"By combining Viz One's proven reliability with intelligent automation, Viz One 8.1 is the next stage of our long-term vision of efficient media operations," said Jochen Bergdolt, Head of MAM at Vizrt. www.vizrt.com
