Speech Recognition with unknown partial feature corruption - a review of the union model

This paper provides a summary of our studies on robust speech recognition based on a new statistical approach – the probabilistic union model. We consider speech recognition given that part of the acoustic features may be corrupted by noise. The union model is a method for basing the recognition on the clean part of the features, thereby reducing the effect of the noise on recognition. To this end, the union model is similar to the missing feature method. However, the two methods achieve this end through different routes. The missing feature method usually requires the identity of the noisy data for noise removal, while the union model combines the local features based on the union of random events, to reduce the dependence of the model on information about the noise. We previously investigated the applications of the union model to speech recognition involving unknown partial corruption in frequency band, in time duration, and in feature streams. Additionally, a combination of the union model with conventional noise-reduction techniques was studied, as a means of dealing with a mixture of known or trainable noise and unknown unexpected noise. In this paper, a unified review, in the context of dealing with unknown partial feature corruption, is provided into each of these applications, giving the appropriate theory and implementation algorithms, along with an experimental evaluation.

Molecular dynamics studies of trinucleotide repeat DNA involved in neurodegenerative disorders.

Expansion of trinucleotide repeat DNA of the classes CAG�·CTG, CGG�·CCG and GAA�·TTC are found to be associated with several neurodegenerative disorders. Different mechanisms have been attributed to the expansion of triplets, mainly involving the formation of alternate secondary structures by such repeats. This paper reports the molecular dynamics simulation of triplet repeat DNA sequences to study the basic structural features of DNA that are responsible for the formation of structures such as hairpins and slip-strand DNA leading to expansion. All the triplet repeat sequences studied were found to be more flexible compared to the control sequence unassociated with disease. Moreover, flexibility was found to be in the order CAG�·CTG > CGG�·CCG = GAA�·TTC, the highly flexible CAG�·CTG repeat being the most common cause of neurodegenerative disorders. In another simulation, a single G�·C to T�·A mutation at the 9th position of the CAG�·CTG repeat exhibited a reduction in bending compared to the pure 15-mer CAGâ�¢CTG repeat. EPM1 dodecamer repeat associated with the pathogenesis of progressive myoclonus epilepsy was also simulated and showed flexible nature suggesting a similar expansion mechanism.