Bringing AI tools to the editors & journalists at IMG Media
IMG Norway was the first customer in the Nordics to use Mimir as part of their daily media management, in March 2019. IMG Sweden quickly followed up, for additional Mimir licenses.
Mimir is a cloud-based media management tool for journalists and video editors, created by Mjoll. With Mimir, you can move your media assets to the cloud and at the same time automate time consuming tasks, such as video and images metadata logging, speech-to-text transcriptions, multi language translations, event tagging, face recognition, and more.
Major requirement for IMG Norway was to have a solution where everyone has access to the content from anywhere. They were on lookout for a solution to gain experience with bringing AI services to end-users, without having to go “all-in” on a specific AI-provider (Google, AWS, IBM, etc).
Mimir was their solution of choice and after only a couple of weeks of using the system, the IMG team reported the following:
“Each of our video editors saves on average 1 day of work a week”
“We see that we can move away from the proprietary systems to cloud based and more lightweight systems even for high-demand Sports productions”.
Trond Larsen, Head of Technical Operations, IMG Norway
THIS IS WHAT MIMIR IS ALL ABOUT!
Save valuable time by automating a number of tasks and give your journalists and editors the tool they need to find relevant videos and images.
IMG MEDIA CHOSE MIMIR AS THEIR MEDIA MANAGEMENT SOLUTION FOR:
- Auto-transcribe to a set of languages
- Automatic metadata logging
- Move archive to the cloud, dating 1-2 years back
HOW IMG USES MIMIR
MULTI LANGUAGE AUDIO TRACKS
At IMG Sweden, they use Mimir to upload commentators clips to cloud storage and auto-transcribe each audio track separately or combined, in the different languages that they produce content in (Norwegian, Swedish, Danish, English). The editors choose a catalog when uploading clips and their catalog name triggers rules for transcribing to correct language.
PERSONS DETECTION
For their Horse Trotting productions, IMG have built their own persons database and added the jockey profiles into Mimir. The jockeys are identified when videos are uploaded and metadata are logged automatically. This is done using AI technology integrated into Mimir. For the persons recognition, AWS Transcribe is used to detect the persons and for logging this as metadata for each of the media assets.
AI TECHNOLOGIES SUPPORTED
With using AI services to analyse videos and images, the editors at IMG Norway save both time and money by automating their video and images metadata logging, speech-to-text transcription, objects detection, and more. All of this, in addition to detecting the Horse Trotting drivers (jockeys) from their own database.
ADVANCED SEARCH IN MIMIR
The editors at IMG can find back to any video clip or images from the easy-to-use Mimir web interface. They can search for persons, any spoken words in any of the languages the clips are transcribed to, in addition to searching for objects, events or any other logged metadata.
INTEGRATED ADOBE WORKFLOW
Mimir integrates with Adobe Premiere Pro, where users can choose to work in a proxy file workflow or a high res workflow. In Mimir, users can search for clips using automatically extracted metadata, including spoken words. Clips can be imported into Adobe Premiere Pro either in its entirety or by selecting text from the transcription which will be turned into subclips that can be imported.
The IMG users can access Mimir from a Adobe Premiere panel for easy access to their videos and images.
This casestudy was originally published on the Mjoll website here.
Contact us about Mimir
Mediability is the reseller of Mimir in the Nordics. We are happy to discuss how Mimir can get you started with moving your media assets to the cloud and with using AI as part of your daily workflow.
sales@mediability.com
+47 482 81 000 (NO)
+46 (0)8 564 85 330 (SE)
+45 33 60 50 60 (DK)