Deep Learning
Efficient Probabilistic Inference in the Quest for Physics Beyond the Standard Model
There are two broad strategies for this type of likelihood-free inference problem.
Positional Normalization
Boyi Li, Felix Wu, Kilian Q. Weinberger, Serge Belongie
A popular method to reduce the training time of deep neural networks is to normalize activations at each layer.
Effective End-to-end Unsupervised Outlier Detection via Inlier Priority of Discriminative Network
Siqi Wang, Yijie Zeng, Xinwang Liu, En Zhu, Jianping Yin, Chuanfu Xu, Marius Kloft
Finally, based on inlier priority, we propose the negative entropy based score as a simple and effective outlierness measure.
HyNet: Learning Local Descriptor with Hybrid Similarity Measure and Triplet Loss
SOSNet [47] shows that second-order constraints further improve the descriptors.
Efficient Forward Architecture Search
Hanzhang Hu, John Langford, Rich Caruana, Saurajit Mukherjee, Eric J. Horvitz, Debadeepta Dey
Neural Information Processing Systems http://nips.cc/
Export Reviews, Discussions, Author Feedback and Meta-Reviews
The left-most column consists of a random sample of hand-written digits from the test set. The other columns are analogies generated from the learned model p(x|z,y). The writing style ('z') of each generated row is inferred from the test-set image in the left-most column, while the class label ('y') iterates from [0,...,9] from left to right. We think this success is quite exciting, and that it accentuates a big advantage of the fully probabilistic approach: the ability to perform meaningful inferences on which the model was not explicitly trained.
Semi-supervised Learning with Deep Generative Models
Durk P. Kingma, Shakir Mohamed, Danilo Jimenez Rezende, Max Welling
Neural Information Processing Systems http://nips.cc/
Articulated Pose Estimation by a Graphical Model with Image Dependent Pairwise Relations
We present a method for estimating articulated human pose from a single static image based on a graphical model with novel pairwise relations that make adaptive use of local image measurements. More precisely, we specify a graphical model for human pose which exploits the fact the local image measurements can be used both to detect parts (or joints) and also to predict the spatial relationships between them (Image Dependent Pairwise Relations). These spatial relationships are represented by a mixture model. We use Deep Convolutional Neural Networks (DCNNs) to learn conditional probabilities for the presence of parts and their spatial relationships within image patches. Hence our model combines the representational flexibility of graphical models with the efficiency and statistical power of DCNNs. Our method significantly outperforms the state of the art methods on the LSP and FLIC datasets and also performs very well on the Buffy dataset without any training.