DeepSat - A Learning framework for Satellite Imagery — mentioned by benanne

Mon, 21 Sep 2015 20:11:39 GMT

I have some doubts about this. Deep learning moves fast and DBNs are pretty much outdated models, even for unsupervised pre-training. It doesn't make much sense to me that unsupervised pre-training would help for this problem to begin with, seeing as their dataset totals around 65TB. The paper is worth checking out: http://arxiv.org/abs/1509.03602 I haven't read it in full, but based on a quick skim, the convnet architectures they evaluated seem laughably tiny and shallow (at most three convolutional layers) by today's standards -- although I appreciate that there may be other constraints at…

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

DeepSat - A Learning framework for Satellite Imagery — mentioned by Thrymr

Mon, 21 Sep 2015 19:16:09 GMT

There's a link to the paper in the article: http://arxiv.org/abs/1509.03602

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

DeepSat - A Learning framework for Satellite Imagery · HackerPapers discussions

DeepSat - A Learning framework for Satellite Imagery — mentioned by benanne

DeepSat - A Learning framework for Satellite Imagery — mentioned by Thrymr