In this paper, a Web-based e-Training platform that is dedicated to multimodal breast imaging is ... more In this paper, a Web-based e-Training platform that is dedicated to multimodal breast imaging is presented. The assets of this platform are summarised in (i) the efficient representation of the curriculum flow that will permit efficient training; (ii) efficient tagging of multimodal content appropriate for the completion of realistic cases and (iii) ubiquitous accessibility and platform independence via a web-based approach.
This paper presents the HealthSign project, which deals with the problem of sign language recogni... more This paper presents the HealthSign project, which deals with the problem of sign language recognition with focus on medical interac- tion scenarios. The deaf user will be able to communicate in his native sign language with a physician. The continuous signs will be translated to text and presented to the physician. Similarly, the speech will be recognized and presented as text to the deaf users. Two alternative versions of the system will be developed, one doing the recognition on a server, and another one doing the recognition on a mobile device.
IEEE transactions on neural networks and learning systems, 2015
In this paper, we propose a Gaussian process (GP) model for analysis of nonlinear time series. Fo... more In this paper, we propose a Gaussian process (GP) model for analysis of nonlinear time series. Formulation of our model is based on the consideration that the observed data are functions of latent variables, with the associated mapping between observations and latent representations modeled through GP priors. In addition, to capture the temporal dynamics in the modeled data, we assume that subsequent latent representations depend on each other on the basis of a hidden Markov prior imposed over them. Derivation of our model is performed by marginalizing out the model parameters in closed form using GP priors for observation mappings, and appropriate stick-breaking priors for the latent variable (Markovian) dynamics. This way, we eventually obtain a nonparametric Bayesian model for dynamical systems that accounts for uncertainty in the modeled data. We provide efficient inference algorithms for our model on the basis of a truncated variational Bayesian approximation. We demonstrate th...
ABSTRACT Hidden Markov models (HMMs) are a popular approach for modeling sequential data, typical... more ABSTRACT Hidden Markov models (HMMs) are a popular approach for modeling sequential data, typically based on the assumption of a first-order Markov chain. In other words, only one-step back dependencies are modeled which is a rather unrealistic assumption in most applications. In this paper, we propose a method for postulating HMMs with approximately infinitely-long time-dependencies. Our approach considers the whole history of model states in the postulated dependen-cies, by making use of a recently proposed nonparametric Bayesian method for modeling label sequences with infinitely-long time dependencies, namely the se-quence memoizer. We manage to derive training and inference algorithms for our model with computational costs identical to simple first-order HMMs, despite its entailed infinitely-long time-dependencies, by employing a mean-field-like ap-proximation. The efficacy of our proposed model is experimentally demonstrated.
In this paper, a Web-based e-Training platform that is dedicated to multimodal breast imaging is ... more In this paper, a Web-based e-Training platform that is dedicated to multimodal breast imaging is presented. The assets of this platform are summarised in (i) the efficient representation of the curriculum flow that will permit efficient training; (ii) efficient tagging of multimodal content appropriate for the completion of realistic cases and (iii) ubiquitous accessibility and platform independence via a web-based approach.
This paper presents the HealthSign project, which deals with the problem of sign language recogni... more This paper presents the HealthSign project, which deals with the problem of sign language recognition with focus on medical interac- tion scenarios. The deaf user will be able to communicate in his native sign language with a physician. The continuous signs will be translated to text and presented to the physician. Similarly, the speech will be recognized and presented as text to the deaf users. Two alternative versions of the system will be developed, one doing the recognition on a server, and another one doing the recognition on a mobile device.
IEEE transactions on neural networks and learning systems, 2015
In this paper, we propose a Gaussian process (GP) model for analysis of nonlinear time series. Fo... more In this paper, we propose a Gaussian process (GP) model for analysis of nonlinear time series. Formulation of our model is based on the consideration that the observed data are functions of latent variables, with the associated mapping between observations and latent representations modeled through GP priors. In addition, to capture the temporal dynamics in the modeled data, we assume that subsequent latent representations depend on each other on the basis of a hidden Markov prior imposed over them. Derivation of our model is performed by marginalizing out the model parameters in closed form using GP priors for observation mappings, and appropriate stick-breaking priors for the latent variable (Markovian) dynamics. This way, we eventually obtain a nonparametric Bayesian model for dynamical systems that accounts for uncertainty in the modeled data. We provide efficient inference algorithms for our model on the basis of a truncated variational Bayesian approximation. We demonstrate th...
ABSTRACT Hidden Markov models (HMMs) are a popular approach for modeling sequential data, typical... more ABSTRACT Hidden Markov models (HMMs) are a popular approach for modeling sequential data, typically based on the assumption of a first-order Markov chain. In other words, only one-step back dependencies are modeled which is a rather unrealistic assumption in most applications. In this paper, we propose a method for postulating HMMs with approximately infinitely-long time-dependencies. Our approach considers the whole history of model states in the postulated dependen-cies, by making use of a recently proposed nonparametric Bayesian method for modeling label sequences with infinitely-long time dependencies, namely the se-quence memoizer. We manage to derive training and inference algorithms for our model with computational costs identical to simple first-order HMMs, despite its entailed infinitely-long time-dependencies, by employing a mean-field-like ap-proximation. The efficacy of our proposed model is experimentally demonstrated.
Uploads
Papers by Dimitrios Kosmopoulos