Hicham Badri

speaker
223 appearances 2 recordings 1 series first heard Jun 2026 last heard 14 Jul

Hicham Badri’s voice in public audio — every appearance, attributed to the second.

Trend

recordings per month · last 12 months
1 · Jul OctJan 26AprJulnow

Recordings per month over the last 12 months — 2 in all, peaking in Jul 2026 with 1.

Appearances

newest first · ▶ plays the moment
Like a bunch of models are running and trying to make sense of the data, etcetera.
Then they were gonna put it put the chest of it, let's say.
Yeah, it's gonna be analyzed and it's gonna be, let's say, summarized.
Then because that part is gonna take a lot of
Work.
So this is gonna be done.
It's gonna like read the audio, understand the transcript, look at the video, understand all these segments and what is going on in the video.
Then it's gonna make like a kind of summary of this.
So when you go and there you search, it's not going and analyzing the whole video.
It's basically just trying to search through this gist or summary.
Let's say uh a question like, Is there a dog in this video, for example?
The model will basically if you're chatting with it, it will
basically connect that that query to what it finds the video, it will take that text and try to search through this summary to find the right segment, if that's possible or not.
If there's actually a dog, we will say, for example, oh, in this video I see a dog around this timestamp.
Like it's a golden retrieval, blah, blah, blah, blah.
Um I mean right now the models are really, really good at detecting and recognizing dogs actually.
Yeah, I think they are even better than humans doing this.
So basically, two steps.
The first step is this pre-processing, where we basically cache and make this kind of uh summary of the video where we know like temporarily like what's going on.
Once the video processed, someone comes, asks a question, we map the question to that summary with a model.
Showing 101–120 of 223 · page 6 of 12 ← Previous Next →