Revolution in video analysis: Moments Lab brings MXT-2!

Revolution in video analysis: Moments Lab brings MXT-2!
On May 7, 2025, Bayer 04 Leverkusen managed to significantly increase efficiency in media production by implementing the AI-based video and photo indexing solution from Moments Lab. The technology automatically recognizes faces, objects, logos, landmarks, texts and audio content and is used by various sports organizations, media companies and journalists to optimize their workflows.
It is particularly noteworthy that Bayer 04 Leverkusen remained unbeaten last season and set a European record of 51 games. The association's social media team regularly shared highlights and entertaining content, for which an average of three and a half hours of film material were recorded every week. Until the introduction of the new technology, the team members had to take notes manually and look through the recordings to cut out the best moments.use of moments lab
The decision to integrate Moments Lab into the production process was welcomed by Robin Klütsch, Senior Visual Designer at Bayer 04. He reported considerable savings from the new solution. The head of innovation, Alexander Hartmann, emphasized how important it is to find relevant content quickly to keep up with the speed of digital marketing.
Moments Lab also recently presented the new MXT-2 AI model on the NAB 2025. This model has a training data set of 1.5 billion pictures, which is more than three times as much as its predecessor. The improvement of the video analysis and the creation of detailed, customizable descriptions are a central goal of MXT-2. The previous procedures in which metadata such as titles and descriptions were added required a lot of time and were difficult to scal.
With MXT-2, customers can find content up to seven times faster from their video archives. The new functions "Custom Moments" and "Custom Insights" offer time -coded descriptions and automatic classification of video content. It should be emphasized that MXT-2 Google Researchs Vid2seq exceeds by 47 % when video sequences.
innovations in video indexing
The "Custom Moments" function allows precise segmentation of content, which includes, for example, the automatic highlighting of dishes in cooking shows or the compilation of highlights in sports reporting. As part of “Custom Insights”, categorization is made possible according to specific parameters, which allows an efficient creation of texts from video content.
With the introduction of MXT-2, Moments Lab sees itself in the position of simplifying the search and sorting of media libraries in the long term. The platform is either available as a video discovery platform or can be integrated into existing tools via an API to accelerate production processes and improve storytelling.
Details | |
---|---|
Quellen |