YouTube viewers will be made privy to ambient sound sub-titles such as [Music], [Applause], and [Laughter].
Adding context for the hearing impaired
Google has announced the addition of the three ambient audio markers, citing their use to enhance a narrative beyond what is simply being said.
The addition of sound effect information to their auto-caption system broadens the viewing experience for YouTube audiences in need of the cues.
Big science behind the ambience
The process is credited to Google’s Deep Neural Network, allowing for more accurate pinpointing of peripheral non-speech audio.
The DNN model currently focuses on three simple categories with plans to incorporate more specific noise notations such as [Ring], [Bark], or [Knock], as well as more accurate musical descriptions of the kind of musical instrumentation being played, as well as the caliber of applause.
These small additions are a big step towards more inclusive captioning and tailoring accessibility to a broader audience.