News for hackers. Stuff that matters.
Re:Show HN: Visual Search using features extracted from Tens... by visualsearchsv on Thursday January 28, 2016 @11:12PM (#10993444)
According to their KDD paper, they showed modified / trained alexnet performing as good as VGG with significantly less computation time. Not sure what they actually use in practice.

http://www.kevinjing.com/visual_search_at_pinterest.pdf

Re:Show HN: Visual Search using features extracted from Tens... by nl on Thursday January 28, 2016 @11:33PM (#10993516)
That sounds like it makes sense, right?

I beleive that most of the strength of VGG (and Inception) vs Alexnet is that it was able to learn the feature relationships better, not because it learnt better features.

VGG is pretty computational intensive, which is why Google concentrated so much on the computational complexity for GoogLeNet/Inception/ReCeption

So if you are just using the features directly then it would make sense to use whatever was the quickest to compute.

Re:Show HN: Visual Search using features extracted from Tens... by visualsearchsv on Thursday January 28, 2016 @11:45PM (#10993553)
Yes, being able to compute quickly is especially important in reducing query latency, much more so than during indexing. What stood out for me in the paper was that out of box performance of VGG (trained on imagenet alone) was as good as fine tuned alexnet.

I am interested in assessing if there are any tricks that could be used when querying from a mobile device. In such cases feature extraction can be performed on the device itself, with only feature vectors sent over the network. In case of pinterest, another special case is that a lot queries are performed on images already present in the system. The user simply readjusts the bounding box to highlight the object of interest. In this case they can simply pre-compute 4~20 crops per image. Online feature computation is much more expensive / complicated than offline.

Re:Show HN: Visual Search using features extracted from Tens... by nl on Friday January 29, 2016 @01:27AM (#10993826)
TensorFlow runs on Android, right? And AlexNet runs on an RaspberryPi, so it should be fine on a phone.

But it would be interesting to know if that is better. I'd imagine most phones have some kind of hardware support for resizing images, so it might be better to take advantage of that and then do feature extraction on a server?

Re:Show HN: Visual Search using features extracted from Tens... by krasin on Friday January 29, 2016 @04:06AM (#10994272)
Contemporary phones (e.g. iPhone 6S) are capable of running GoogLeNet at 1 FPS, see, for example, this demo (mine): https://github.com/krasin/MetalDetector

AlexNet will run at ~10 FPS, I guess.