Visual Explorer: Automatic Transcription & Translation of Iran's IRINN Now Live

In collaboration with the Internet Archive's TV News Archive, we are tremendously excited to announce that we are now live-transcribing and translating into English all coverage from Persian-language Iranian state television channel IRINN from January 29th to present. Immediately upon the Archive completing processing of a given broadcast, we transcribe it using Google's Cloud Speech-to-Text API and then translate into English using Google's Cloud Translation API using a unique workflow that passes timecode information transparently through the translation pipeline.

The end result is that all IRINN broadcasts monitored by the Archive will appear in the Visual Explorer within a few hours of broadcast: first in a Persian-language machine transcript and then shortly after that in machine-translated English.

We are incredibly excited about the profound new opportunities this makes possible. While imperfect, machine transcription and translation removes the language barrier to at least understanding the general gist of television news from across the world, making it possible for journalists, scholars and mis/disinformation and propaganda experts to understand the realtime narratives of the Iranian state regarding everything from global events to its crackdown on domestic protests for women's rights.

We'd love to hear from you with any questions about this collection!

Launch Visual Explorer.