News for hackers. Stuff that matters.
Re:Extracting image metadata at scale by zetazzed on Monday March 21, 2016 @03:43PM (#11331219)
The authors may want to consider how much of this work could be done easily and effectively with deep learning. For content-based search and image similarity, even simple, pre-trained convnets will likely crush the histogram-based approaches you have here.

Just run your images through Google Cloud vision to do the face detection and text detection. With 2M images, it will be cheaper than the amount of dev time you spent here, and you'll get excellent quality.

Re:Extracting image metadata at scale by TTPrograms on Monday March 21, 2016 @03:53PM (#11331305)
They explain that not all of the images they want are faces in this case, so you'd have to train your own on "interesting regions" (though there is some work in that area). Part of the challenge in that case is generating all your labels for what the interesting regions are. This way they don't need to generate labels, at least.
Re:Extracting image metadata at scale by aab0 on Monday March 21, 2016 @07:30PM (#11333094)
Youtube did something similar for 'interesting thumbnails' last year with deep nets (many uploaders do not specify a good thumbnail preview), and reported that it gave a nice performance boost.
Re:Extracting image metadata at scale by shekkizh on Monday March 21, 2016 @04:06PM (#11331426)
My thoughts as well. With the recent advances doing image similarity and content based search should be fairly simple and probably more effective with pre-trained convnets.
Re:Extracting image metadata at scale by maciejgryka on Monday March 21, 2016 @03:49PM (#11331274)
It would be awesome if they released this data for another edition of the Netflix challenge, Then we could all try this ourselves!