Ross Girshick @inkynumbers

Giver of vision to machines. rossgirshick.info Joined April 2009

Tweets

28
Followers

4K
Following

42
Likes

14

Ross Girshick @inkynumbers

a year ago

Do you want to work with me and our amazing team? Apply at vercept.com/careers Stack: PyTorch, React, Next.js, TypeScript #PyTorch #NextJS #AI #React #AIStartup

7 15 128 36K 53

View Details

@giffmana FAIR was and continues to be an amazing place! However, being in one place for a long time (8 yrs for me) can eventually become its own good reason to move. Reinitialization and randomization are important in research life. (Any talk of publication quotas is pure nonsense.)

3 7 213 30K 7

View Details

Ross Girshick @inkynumbers

4 years ago

@gruntleme It's a wonderful illustration of a giant house spider! These friends visit me from time to time in the basement (where I work from home). They have a curious range: most of Europe, a small bit of the PNW around Seattle and Vancouver, BC, and another small bit of the Mid-Atlantic.

0 0 0 0 0

View Details

Ross Girshick @inkynumbers

5 years ago

@karpathy @giffmana @PaulKRubenstein @endernewton @sainingxie The mismatch may be an issue (I don't know), but apparently it's not a catastrophe. End-to-end or partial fine-tuning may help compensate, if it is a problem. I also find it somewhat concerning and think it could be worth investigating.

1 0 4 0 0

View Details

Ross Girshick @inkynumbers

5 years ago

@giffmana @PaulKRubenstein @endernewton @sainingxie Yes, all patches are used for downstream tasks.

1 0 7 0 0

View Details

Ross Girshick @inkynumbers

5 years ago

@giffmana I agree about the use of nparams being slightly wrong here. Within the scope of the fig each "column" of points is comparable, but using flops on the x-axis would be more meaningful wrt scaling. (I actually complain about nparams as a complexity measure all the time...)

0 0 0 0 0

View Details

Ross Girshick @inkynumbers

5 years ago

@giffmana @y_m_asano @endernewton @sainingxie Our bias is for det and seg transfer, rather than more cls results, so that's what we went with (lacking bandwidth for both). I do want to note the det and seg tables show IN1k sup baselines (not just self-sup). MAE (and BEiT) surpass IN1k sup convincingly, exciting to me!

1 0 6 0 0

View Details

Ross Girshick @inkynumbers

5 years ago

@giffmana It's a bit buried in the caption of Table 3, but therein it says ViT-H is /14. We debated the line-plot style and have diverging opinions of what looks nice ;).