A novel paper-based system for editing media using transcripts
Project from - present
What we're doing
We have developed a novel paper-based system for editing media using transcripts. Producers can print out a transcript of their recording onto normal paper and use an Anoto digital pen to either underline or cross-out words. The media is then automatically edited to reflect these annotations. This tool allows producers to quickly and efficiently edit their recordings away from the screen and on-the-go.
Why it matters
Many producers already make a ‘paper edit’ of their content by transcribing recordings manually, printing them out and using a pen to mark which bits of the recording they do or don’t want. After this, they must use editing software to find and edit the content, which is a tedious process. Our system provides them with an automatically-generated transcript, and allows them to edit it directly on the page, saving them from having to transcribe the recording themselves or go back to make the edits.
How it works
The paper editor uses a speech-to-text engine to generate transcripts with precise timestamps for each word. The transcript is printed out on normal paper with a special dot pattern in the background. This pattern allows a digital pen to record which bit of the page it’s writing on. After the producer annotates the page and docks the pen, the annotations are uploaded to our system which links the selected words to the timestamps for those words. This generates an edit decision list (EDL) of the edit which can be opened in audio or video editing software.
We have collaborated with digital pen manufacturer Anoto to develop a prototype of the system which is currently being tested with our colleagues in TV and radio.
This project is part of the Immersive and Interactive Content section
This project is part of the Audio Research work stream