Quickly search audio and video evidence with transcription in Magnet Axiom and Axiom Cyber
The ability to quickly capture high-quality video, live and motion photos, and audio clips with phones, tablets, and personal computers has led to a drastic increase in media being created and shared. User-generated audio and video is extremely prevalent in both messaging and social media with popular platform WhatsApp reporting that an average of 7 billion voice messages are sent on the platform daily.
With this volume of media also appearing in forensics investigations, we have added the ability for Axiom and Axiom Cyber to transcribe audio and video files to text, facilitating the search, analysis, and review of media evidence.
Searching transcription to surface evidence
With a transcription of the audio for media clips, searching through the content becomes incredibly fast and efficient. Transcriptions of audio and video evidence in Axiom are searchable using the Global Keyword Search or the Local Search in the transcription card to find specific terms or phrases in the media.
Once an area of interest is identified, you can simply click on the transcription timestamp to set the media player playback, helping to quickly surface both the playback and text for potentially relevant content and providing detailed context on the media.
Reviewing and sharing transcriptions
When reviewing media in full screen mode, the transcription card will detach from the main navigation to display over top of the media player and can be repositioned as needed. The transcribed text associated with the current timestamp is highlighted as the media is played, allowing you to easily navigate through the content. The two-way synchronization of the media playback and text is especially helpful when reviewing the lead-up events following, a time of interest.
The transcription function isn’t limited to a single language and will transcribe audio from multiple languages which can then be translated using the included translation module in Axiom and Axiom Cyber with coverage for 43 languages. Results can also be easily exported and shared with downstream reviewers via Portable case, a lightweight version of the full capabilities found in Axiom—designed for easy access and analysis of forensic findings.
Transcription as a tool
As with any autogenerated content, the transcription should always be verified for accuracy, as there are several factors that can come into play including the audio quality of a recording and terminology used which may not be common vernacular. This is in line with the first point of our Founding Principles of AI at Magnet Forensics: Design AI to enhance—never replace—human judgment.
Axiom provides the ability to include notations where you feel that the transcription requires corrections or added clarity. Or, if the quality of the audio isn’t sufficient to support an accurate transcription, you can remove the transcription for either a selected artifact or all transcriptions in the case.
Getting started with transcriptions
Within Axiom you will be prompted to install the necessary model to support this new AV transcription capability. Once you have installed the model, you will be able to transcribe text from a single or all audio and video artifacts in Axiom Examine. From the Artifact Explorer or Media Explorer, you can access the new transcription capabilities from the preview card for the artifact.
The transcription module is included in Axiom Cyber and our Axiom Advanced and Premier term licenses. If you are already using one of these licenses, simply update to the latest version and check out our support article on setting up Transcription of audio and video in the Magnet Forensics Support Portal. If you aren’t already using Axiom or aren’t on an applicable license, you can request a request a free trial to experience audio and video transcription for yourself.