End-to-End Action Detection with Dynamic Contextual Mapping – We propose a deep-learning method to predict the action using contextual MAPs for action prediction in real-time. The state of the art works use a mixture of the following two strategies. First the action prediction is used to predict whether two events should be considered as a single event and by what order they should be considered. Second the Mapped Contextual Mapping is used to predict whether the action should be considered as a chain event or a sequence of events. Finally the Contextual Mapping for the action prediction is used to predict the sequence of events from the contextual data in order to predict the action’s outcome. Compared to state-of-the-art deep learning methods, our method outperforms them in terms of accuracy and speed.

This paper presents a machine learning method for analyzing non-local and local features to obtain a better understanding in the human expert system. In particular, we present two approaches: a model based on multilayer perceptron (MLP) and a model based on a deep learning approach (DeepNet). The MLP model can be easily designed using image processing and training images to extract local details from the features. The DeepNet model can be easily designed using an image representation model. The MLP model can be learned using an image representation model (RBM) and the DeepNet model can be easily designed using an image representation framework and a deep learning approach. We illustrate our method in terms of training data representation on a dataset of human experts. By training the MLP model on an image representation model, we build a dataset of experts who do not use the MLP model.

Proceedings of the 38th Annual Workshop of the Austrian Machine Learning Association (ÖDAI), 2013

Sparse Representation based Object Detection with Hierarchy Preserving Homology

End-to-End Action Detection with Dynamic Contextual Mapping

Mining for Structured Shallow Activation Functions

Evaluating the quality of lexico-semantic prediction in the medical jargonThis paper presents a machine learning method for analyzing non-local and local features to obtain a better understanding in the human expert system. In particular, we present two approaches: a model based on multilayer perceptron (MLP) and a model based on a deep learning approach (DeepNet). The MLP model can be easily designed using image processing and training images to extract local details from the features. The DeepNet model can be easily designed using an image representation model. The MLP model can be learned using an image representation model (RBM) and the DeepNet model can be easily designed using an image representation framework and a deep learning approach. We illustrate our method in terms of training data representation on a dataset of human experts. By training the MLP model on an image representation model, we build a dataset of experts who do not use the MLP model.