We propose a novel multi-stream framework for continuous conversational speech recognition which employs bidirectional Long Short-Term Memory (BLSTM) networks for phoneme predicti...
— Consistency of object dynamics, which is related to reliable predictability, is an important factor for generating object manipulation motions. This paper proposes a technique ...
This article explores the utility of neural network ensembles in knowledge discovery and integration. A novel neural network ensemble model KBNNE (Knowledge-Based Neural Network E...
Temporal derivatives are computed by a wide variety of neural circuits, but the problem of performing this computation accurately has received little theoretical study. Here we sy...
The solving of least square systems is a useful operation in neurocomputational modeling of learning, pattern matching, and pattern recognition. In these last two cases, the soluti...