We propose a new method for handwritten word-spotting which does not require prior training or gathering examples for querying. More precisely, a model is trained "on the fly...
Abstract. Climate models are complex mathematical models designed by meteorologists, geophysicists, and climate scientists to simulate and predict climate. Given temperature predic...
— Given an AWGN channel, we look at the problem of designing a source controlled binary antipodal signaling system for transmitting blocks of binary symbols generated either by a...
Pedro M. Crespo, Estibaliz Loyo, Javier Del Ser, C...
In this paper, we survey the current state-ofart models for structured learning problems, including Hidden Markov Model (HMM), Conditional Random Fields (CRF), Averaged Perceptron...
For Hidden Markov Models (HMMs) with fully connected transition models, the three fundamental problems of evaluating the likelihood of an observation sequence, estimating an optim...