In this paper, we propose a framework for predicting the performance of a vision algorithm given the input image or video so as to maximize the algorithm's ability to provide...
Abstract. We present an implementation of model-based online reinforcement learning (RL) for continuous domains with deterministic transitions that is specifically designed to achi...
This paper revisits the pivot language approach for machine translation. First, we investigate three different methods for pivot translation. Then we employ a hybrid method combin...
This paper presents a header compression algorithm that unlike previous protocols is capable of compressing MAC headers in a multiple-access (shared) channel. Previous schemes coul...
This paper presents a novel facial expression recognition methodology. In order to classify the expression of a test face to one of seven pre-determined facial expression classes, ...