Accelerating large vocabulary continuous speech recognition on heterogeneous CPU-GPU platforms.
Jungsuk KimIan R. LanePublished in: ICASSP (2014)
Keyphrases
- heterogeneous computing
- graphics processors
- graphics processing units
- gpu implementation
- data transfer
- general purpose
- real time
- parallel implementation
- parallel processing
- grid computing
- parallel computing
- automatic speech recognition
- real world
- compute intensive
- operating system
- information retrieval
- parallel computation
- speech recognition
- data structure
- neural network
- data sets
- memory bandwidth
- database