NAB Show: Moments Lab To Unveil Next-Generation MXT-1.5 AI

AI and cloud media company Moments Lab, formerly Newsbridge, will showcase new features powered by the latest version of its award-winning AI indexing models at the NAB Show in Las Vegas (April 14-17) in Booth SL2113.

“Attendees will see how MXT-1.5’s new sequences detection and sound bites features automatically highlight the best moments in a video to help editors and producers create content faster,” the company said.

Philippe Petitpont, Moments Lab co-founder/CEO, said: “In the past year, MXT-1 has helped many organizations realize the potential of generative and multimodal AI when deployed for various use cases — such as boosting the searchability of their vast media collections.

“Our users are finding the exact moments they need in seconds, without any special media search know-how. As a result, they’ve slashed their content sourcing times by as much as 70%. At this year’s NAB Show, we’re upping the innovation with new AI features for the automatic indexing and summarizing of video content that we know the industry will love.”

New features include:

• Sequences Detection | Enabling users to instantly find the moments they need to build a story, MXT-1.5 improves the relevancy of search results by automatically grouping shots into the editorial sequences that content producers are looking for.
• Automatic Sound Bites | MXT-1.5 automatically highlights soundbites in a video, allowing journalists and editors to instantly discover compelling quotes to help tell their stories without needing to spend time scrolling an entire video or transcript.
• Media Topic Categorization |MXT-1.5 automates the categorization of videos into relevant overarching topics or themes defined by the user. By handling the mundane, time-consuming task of content classification, the technology enables media managers to focus on more strategic and creative tasks.

Also, Moments Labs says, it will unveil “powerful new features in its most sophisticated AI indexing models yet. Video scene descriptions generated by MXT-1.5 are 87.5% more accurate than other open source vision language models. This unlocks significant efficiencies for media asset indexing. Using MXT-1.5, content producers can bring up the most relevant moments in their extensive media libraries in seconds. MXT-1.5’s new features and improvements ultimately empower media organizations to create more compelling and monetizable content, faster.”