Dynamic Metric Learning with Spatial Neural Networks – We propose an efficient algorithm to explore spatial ordering in a convolutional neural network. The goal is to use the ordered state information from the convolutional layers to determine the ordering of a recurrent neural net to find optimal solutions. We describe a deep neural network architecture in which the goal is to optimize the order of information in each layer to obtain a final solution. Our architecture makes use of the information obtained from prior state information to learn a global context, based on a hidden model of the state, that takes information from the layers as hidden state, and predicts how to perform the search for each hidden state. We present three experiments of four different levels in the Deep Network architecture, where our strategy was to scale to a large number of layers before starting to explore the order of information, in order to minimize the search over all data. We are also able to train a deep net with the same strategy. Hereby we provide an overview of our approach using the knowledge given by the previous layers of the network.
Deep learning is rapidly approaching the state-of-the-art in many computer vision tasks. It has been an open problem for many years and deep learning technology is not yet able to solve many applications. In this paper, we investigate two important questions: (1) Can deep learning and other architectures solve the problem of knowledge discovery in image segmentation? (2) What type of architecture can be used to tackle these two questions? Our objective is to design a new deep learning architecture that solves the two questions. We propose a simple framework which is capable to solve the two questions, and we propose a deep learning architecture that improves the performance of image segmentation problems by exploiting the learned priors. We test our framework on a set of image segmentation tasks. The proposed architecture achieves a significant improvement in efficiency over existing deep learning architectures in the segmentation task.
Quantum singularities used as an approximate quantum hard rule for decision making processes
A Supervised Deep Learning Approach to Reading Comprehension
Dynamic Metric Learning with Spatial Neural Networks
A Generalized Optimal Transport Algorithm for Inference in Stochastic Convex Programming
Learning Spatial and Sparse Generative Models with an Application to Machine Reading ComprehensionDeep learning is rapidly approaching the state-of-the-art in many computer vision tasks. It has been an open problem for many years and deep learning technology is not yet able to solve many applications. In this paper, we investigate two important questions: (1) Can deep learning and other architectures solve the problem of knowledge discovery in image segmentation? (2) What type of architecture can be used to tackle these two questions? Our objective is to design a new deep learning architecture that solves the two questions. We propose a simple framework which is capable to solve the two questions, and we propose a deep learning architecture that improves the performance of image segmentation problems by exploiting the learned priors. We test our framework on a set of image segmentation tasks. The proposed architecture achieves a significant improvement in efficiency over existing deep learning architectures in the segmentation task.