Most Semantic Web data visualization tools structure the representation according to the concept definitions and interrelations that constitute the ontology’s vocabulary. Insta...
Benjamin Bach, Emmanuel Pietriga, Ilaria Liccardi,...
We show how a preselection of hidden variables can be used to efficiently train generative models with binary hidden variables. The approach is based on Expectation Maximization (...
The aim of latent semantic indexing (LSI) is to uncover the relationships between terms, hidden concepts, and documents. LSI uses the matrix factorization technique known as singu...
Multiple pitch estimation consists of estimating the fundamental frequencies and saliences of pitched sounds over short time frames of an audio signal. This task forms the basis of...
We present an algorithm for dereverberation of speech signals for automatic speech recognition (ASR) applications. Often ASR systems are presented with speech that has been record...
Kshitiz Kumar, Rita Singh, Bhiksha Raj, Richard M....