VIDBE-Q 2025 Volume 70 Issue 2
American Sign Language (ASL) interpretation. Currently, the media accessibility
tool is in the beta testing phase.
The AI-assisted system allows users to upload a video and then choose
which accessibility feature they would like to start with. It is recommended to
begin with captions. Having captions that are known to be accurate helps in the
creation of audio description and ASL later on. Users can choose either English or
Spanish as the source language, then the system creates a "first pass" of the
captions using AI models and natural language processing. This takes only a few
seconds to a few minutes, depending on the length of the video. Typically, the
caption text has been surprisingly accurate if the original audio is very clear.
However, it is strongly recommended that the captions be reviewed and edited
before final approval. Also, it must be noted that speaker identification and sound
effects are not yet automatic. For now, they must be added manually. Another
unique feature is that the DCMP Captioning Key guidelines for line breaks and
caption segmentation are implemented automatically. This is a crucial aspect of
readability that no other automatic captioning system includes. The system also
allows the user to choose the intended grade level of the user, and this sets the
presentation rate of the captions. If there are areas where the captions are too fast,
this will be displayed in red to warn the user that an adjustment is necessary.
Exports currently include .vtt and .srt; other file types can be added upon request.
