On-screen text from video extracted/included in transcript
J
Jason
On screen text is very important for context and content creation (ex. Written hooks). Currently Recall supports text within images but not video, to my knowledge. It would be very helpful and at some point necessary for on screen text to be extracted, included in time stamped summary, and searchable as well.
Log In
s
schroell
Indeed very funny. In some internal channels I already have suggested exactly the same thing. Here is my request text: BavariaR [A\], — 9/1/26, 8:12 PM
Another idea or use case came to my mind, especially when it comes to YouTube videos. Some YouTube videos have not only an audio transcript that can be easily captured, but also slides with content. It would be extremely cool if, when sending to Recall, you could switch atoggle and ask Recall to not only transcribe the spoken text in the video but also analyze the video for interesting content. That content would need to be OCR'd with a visual AI LLM model. I think you get the point.
J
Jason
schroell there’s a similar app called second brain that does this I believe. I want it to analyze videos for content creation to identify visual hooks, written hooks, camera angle or scene change and so what you’re talking about is a yes too. Maybe not ever video but if we click for a deeper analysis yes.
P
Paul Conway
Text-to-speech reading is also important for visually challenged and dual cognitive processing learners.
J
Jason
Paul Conway I think it would do that already if that text is added into the transcript/summary.