Fei-Fei Li
speaker
562 appearances
2 recordings
1 series
first heard Nov 2025
last heard 25 Nov
Fei-Fei Li’s voice in public audio — every appearance, attributed to the second.
Trend
recordings per month · last 12 monthsRecordings per month over the last 12 months — 2 in all, peaking in Nov 2025 with 2.
Appearances
So I was talking to Justin and Andrea about
Can you go what we what we end up calling dense captioning, which is, you know, uh describe the scene in greater details, especially different parts of the scene.
So that's
only I was doing it in one pass, uh y you also optimize inference.
Uh you're doing it on a webcam, I remember.
Yeah, yeah.
Yeah.
it was uh it was pretty impressive. 'Cause most of my graduate students would be satisfied if they can publish the paper, right?
They they package the the the research, put it in a paper, but Justin went a step further.
He's like, I wanna do this real time web demo.
I just wrote a whole manifesto on
I think they are different.
Um I do think the architecture uh of uh these generative models will share a lot of uh shareable components, but uh I think the deeply
Three D four D spatial world has a level of
of structure that is fundamentally different from a purely generative uh signal that is one dimensional.
Well you can also think it's sound, but even sound even sound you can translate into corelog yeah, you get corelogram which is a two D signal.
But there's no indication that th those latent uh uh modeling will get you to a causal law of uh of space and dynamics, right?
That's where today's deep learning and uh
human intelligence actually start to bifurcate because fundamentally the deep learning is still fitting patterns.
If you use the word understand the way you understand, I'm pretty sure the model doesn't understand it.
Showing 101–120 of 562 · page 6 of 29
← Previous
Next →