dumitrue
Joined 14 years ago
Karma
440
Submissions
38
Member since
Jul 30
Recent submissions
Comment · 10 years ago
Apart from the captioning work already mentioned, there's also visual analogy making work by Reed et al. http://www-personal.umich.edu/~reedscot/nips2015.pdf which is pretty exciting (sort of a converse to what you're proposing).
Comment · 11 years ago
But the test set has 50,000 images (across 1000 categories). It's not that easy to just "try many times" to get it right.
Comment · 11 years ago
I don't think OpenCV really solved computer vision to be fair. There's definitely no model out there that can do image-based question & answering as well as a human can, or interpret the contents of an image (parse it, if you will) in an accurate way, with the exception of very few special cases.
Comment · 11 years ago
Not really -- he's been an active Google+ poster even before he joined FB (https://plus.google.com/+YannLeCunPhD/posts)
Comment · 12 years ago
Just went out: http://arxiv.org/abs/1411.4555
Comment · 12 years ago
Just FYI, the only additional data used by the GoogLeNet entry was from the classification challenge (aka provided by the organizers), hardly something that would make you lose sleep at night.
Comment · 13 years ago
The Los Angeles office is right by Venice Beach, in a beautiful setting. It's a mid-sized office (~500 employees according to http://venice.patch.com/groups/business-news/p/silicon-beach...) in the same time-zone as Mountain View and a less than 1h flight to the mothership should you need to go there.
Unlike the main office, most people don't have to choose between commuting from SF or living in Mountain View, because Venice/Santa Monica is actually a nice area to live in :). Naturally, the breadth of projects is not as big as in Mountain View, but there's a number of exciting things happening here (computer vision, quantum AI etc).
Comment · 13 years ago
Yes, though I think on public benchmarks this is still not the case. There's a dog-breed classification problem in this year's Fine-Grained challenge (https://sites.google.com/site/fgcomp2013/) so we'll see in December!
Comment · 13 years ago
It's actually already possible to train convolutional network-like models to distinguish between a variety of dogs, cats etc with precision that is pretty much super human. The real problem is getting high-quality training data without involving tons of domain experts that would tells us with high degree of confidence whether a given image is of a specific breed of dog (getting millions of images of dogs is easy, so is building a classifier).
It's not immediately obvious to me how useful such an app would be btw. Unless I of course misunderstood what a "real life pokedex app" is :).
Comment · 13 years ago
Yes, it'd be surprised if the straightforward implementation from https://code.google.com/p/cuda-convnet/, run on a GPU with lots of transformations, wasn't the winning entry.
http://research.google.com/pubs/DumitruErhan.html
http://www.researchgate.net/profile/Dumitru_Erhan
http://www.dumitru.ca/