Building high-level features using large scale unsupervised learning · HackerPapers discussions

Building high-level features using large scale unsupervised learning — mentioned by visarga

Wed, 26 Aug 2015 20:49:04 GMT

Nope. Here's the link to the paper Google team published in 2002. Building high-level features using large scale unsupervised learning http://arxiv.org/abs/1112.6209 From the abstract: Contrary to what appears to be a widely-held intuition, our experimental results reveal that it is possible to train a face detector without having to label images as containing a face or not. The Cat detection thing was just a side product of learning to identify features of things in an unsupervised manner, but the news outlets locked on to that with titles such as "How Many Computers to Identify a Cat?…

Paper and discussion history. A mention is not an endorsement.

Building high-level features using large scale unsupervised learning — mentioned by nl

Sun, 15 Dec 2013 22:34:33 GMT

But Google is (historically) a software company, so we should compare its contributions in the software field at the moment. Here's a random list: Page Rank Map/Reduce Statistical Language Translation Unsupervised image feature extraction[1] Go Dart Large scale software defined networking. The list could go on.. [1] http://arxiv.org/abs/1112.6209

Paper and discussion history. A mention is not an endorsement.

Building high-level features using large scale unsupervised learning — mentioned by kdavis

Wed, 25 Sep 2013 16:39:49 GMT

The "state of the art" they reference is SVM's trained on color and texture features. Pre deep belief network I'd agree with your guess on convolutional neural networks. However, now I'd guess you'd use a deep belief network to create a network that would pick out better features than those picked out "by hand" in the convolutional neural network. (See for example [1][2]) So my money would be on some deep belief network. [1] Hinton, G. E, Osindero, S., and Teh, Y. W. (2006). A fast learning algorithm for deep belief nets. Neural Computation, 18:1527-1554. [2] Building high-level features…

Paper and discussion history. A mention is not an endorsement.

Building high-level features using large scale unsupervised learning — mentioned by varelse

Fri, 30 Aug 2013 16:54:59 GMT

Given the 15.8% success rate of the cat detector on large data sets: http://arxiv.org/pdf/1112.6209.pdf&embedded=true, My chief worry is overconfidence from hopelessly technologically ignorant bureaucrats combined with a slew of false negatives and false positives when applied to human faces on general surveillance feeds. In contrast, driver's license, ID, and arrest photos are at least severely constrained input filters wherein a single individual is placed against relatively uncomplicated background pixels and the face is in roughly the same orientation.

Paper and discussion history. A mention is not an endorsement.

Building high-level features using large scale unsupervised learning — mentioned by cs702

Tue, 26 Jun 2012 18:26:37 GMT

The system apparently did not require any training by humans; instead, it learned to catalog different types of images, including those of human faces, on its own. (In other words, all images were unlabeled.) Paper here: http://arxiv.org/abs/1112.6209v3 [UPDATE: fixed erroneous statement.]

Paper and discussion history. A mention is not an endorsement.

Building high-level features using large scale unsupervised learning — mentioned by yk

Tue, 26 Jun 2012 03:52:29 GMT

The original article appears is http://arxiv.org/abs/1112.6209 (as noted in the article). I did only print it (yet), but from the abstract it appears that they did mainly tried face recognition, not trying to find cats.

Paper and discussion history. A mention is not an endorsement.

Building high-level features using large scale unsupervised learning — HN submission

Wed, 30 May 2012 06:37:03 GMT

Google research (with Jeff Dean): Face recognition with unsupervised learning

3 points; 0 discussion replies on HN. Paper and source history. A submission is not an endorsement.