Hoang M. Le bio photo

Hoang M. Le

Staff Research Scientist @ Latitude AI.

Twitter LinkedIn Github e-Mail

This is my customary update of percentage of deep-learning papers in the major vision conferences. At CVPR 2015, I described its exponential growth, after ICCV 2015 I wondered whether it had plateaued, and in CVPR 2016 it took over again.

Here the evolution of the results as of ECCV 2016:

CVPR2013:  0.85% (  4 out of 471)
ICCV2013:  1.54% (  7 out of 455)
CVPR2014:  3.70% ( 20 out of 540)
CVPR2015: 14.45% ( 87 out of 602)
ICCV2015: 14.45% ( 76 out of 526)
CVPR2016: 23.48% (151 out of 643)
ECCV2016: 20.96% ( 87 out of 415)

And the plot using XKCD style, as described in this blog post:

...

So in ICCV15 we had in fact a pit stop which has been surpassed by CVPR16 and ECCV16. Will ECCV16 be also a pit stop or we will be finally saturating?

It is also interesting to note the smaller size of ECCV compared to the rest. I prefer it this way, to be honest:

See you in Amsterdam!