How about adding more advanced models (mobile net, DPN, etc) in vision ?

[Naruto-Sasuke]

Mobile net, DPN, SE-ResNeXt are useful in image classification. I wonder whether you guys have a plan adding these models.

[vfdev-5]

There is also a repository with pretrained recent architecture models which contains DPN, SE-ResNeXt.
HTH

[fmassa]

We have added resnext and mobilenet. I'm closing this issue and concentrating the discussion in #645