#None #AI Video #Moments Lab #MXT-2

Moments Lab Launches MXT-2: Transforming Video Discovery

In a bold move to enhance the capabilities of video content creators, Moments Lab, a prominent AI video discovery firm, has announced the release of its latest technology, the MXT-2. This state-of-the-art multimodal AI indexing system is engineered to revolutionize the way producers interact with large video libraries, making the content creation process faster and more efficient.

The MXT-2 stands out due to its impressive training dataset, which comprises over 1.5 billion images. This is more than three times the size of MXT-1, its predecessor. Thanks to this expansive dataset, MXT-2 is capable of generating human-like video descriptions with astonishing precision and detail. This leap in technology not only increases the accuracy of video indexing but also enhances the searchability of video content, allowing creators to find the perfect clips for their projects quickly.

A significant feature of MXT-2 is its Custom Moments capability. This functionality provides users with time-coded descriptions of essential video segments that cater to individual preferences. For instance, it can help identify the most engaging scenes in unscripted television programs, enabling creators to easily compile viral shorts or trailers. This customization allows for a more focused approach to content creation, fostering creativity and innovation among media professionals.

In addition to Custom Moments, the Custom Insights feature dramatically extends MXT-2’s capabilities. It can automatically generate texts of varying lengths, including titles, teaser descriptions, and comprehensive articles. Furthermore, it classifies videos by theme or editorial topic, streamlining the workflow for content producers and enriching their storytelling processes. “We are proud to claim that MXT-2 is the most advanced technology available for sequencing and describing video content,” remarked Frederic Petitpont, CTO and co-founder of Moments Lab. He emphasized the AI’s superiority by noting that it surpasses Google Research's Vid2Seq in video sequencing accuracy by 47%.

Philippe Petitpont, CEO and co-founder of Moments Lab, expressed his enthusiasm about this launch, stating, “Our mission is to eliminate barriers to effective video storytelling. With MXT-2, we aim to simplify the process of searching and organizing expansive media libraries.” By providing deeper editorial insights and improved video searchability, MXT-2 empowers users to maximize their content creation efforts, reuse materials effectively, and ultimately unlock new revenue streams.

The MXT-2 tool is readily available on Moments Lab's video discovery platform and can also be integrated into existing workflows via API. It will be showcased in a live demonstration at the upcoming 2025 NAB Show, which will take place from April 6-9 in Las Vegas at booth SL12213. Interested attendees will have the opportunity to see firsthand the capabilities of this groundbreaking AI technology.

Moreover, Moments Lab is also making a mark at the NAB Show’s BEIT Conference, where Khalil Guetari, an AI software engineer at the company, will present insights on "Vision and Language Models for Enhanced Archive Video Management" on April 7. This presentation is expected to dive deeper into how AI can facilitate better management of archival video content.

Moments Lab, with its commitment to enhancing video production processes, continues to support a wide array of industries, from television broadcasters to sports rights holders and brands. The MXT-2 is designed to address the challenges associated with video understanding and indexing, enabling professionals to devote more time to creativity and thoughtful decision-making while leaving the nitty-gritty of video analysis to the AI.

To explore more about MXT-2 and its features, please visit Moments Lab's official website.

With technologies like the MXT-2 at the forefront, the future of video content creation looks promising, allowing creators to focus on what truly matters: telling compelling stories through visually stunning media.