Large Vocabulary Speech Recognition System on Parallel Architectures
Speaker: Patrick Cardinal, Centre de Recherche d
Date: Monday, September 17 2012
Time: 4:00PM to 5:00PM
Refreshments: 3:45PM
Location: 32-G882 (Stata Center - Hewlett Room)
Host: Jim Glass, MIT CSAIL
Contact: Marcia Davidson, 617-253-3049, marcia@csail.mit.edu
Relevant URL: The speed of modern processors has remained constant over the last few years but the integration capacity continues to follow Moore's law and thus, to be scalable, applications must be parallelized. In addition to the main CPU, almost every computer is equipped with a Graphics Processors Unit (GPU) which is in essence a specialized parallel processor.
In this presentation, I will discuss the use of GPU for acoustic likelihood computations and how the A* search can be used instead of the classical Viterbi algorithm for improving the performance on multicore processors.
See other events happening in September 2012
