Multi-scale Orderless Pooling of Deep Convolutional Activation Features — mentioned by rsfern

Wed, 08 Jun 2016 02:15:19 GMT

I think [0] or [1] might be a good place to start, depending on what you want to do with it. Both take a similar to this blog post, in that they also use the VGG networks trained on ImageNet. Specifically, [0] uses Fisher Vector pooling on the highest layer of convolution features, and they report good performance on texture and scene datasets even though they used weights learned from ImageNet data. Plus, skipping the fully connected layers means you can in principle use any image size/resolution without worrying about the size of the model. [0]: http://arxiv.org/abs/1411.6836 [1]…

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

Multi-scale Orderless Pooling of Deep Convolutional Activation Features · HackerPapers discussions

Multi-scale Orderless Pooling of Deep Convolutional Activation Features — mentioned by rsfern