| Presentation # | 4 |
| Session: | ASR III (End-to-End) |
| Location: | Kallirhoe Hall |
| Session Time: | Friday, December 21, 10:00 - 12:00 |
| Presentation Time: | Friday, December 21, 10:00 - 12:00 |
| Presentation: |
Poster
|
| Topic: |
Speech recognition and synthesis: |
| Paper Title: |
ON-DEVICE END-TO-END SPEECH RECOGNITION WITH MULTI-STEP PARALLEL RNNS |
| Authors: |
Yoonho Boo, Jinhwan Park, Lukas Lee, Wonyong Sung, Seoul National University, Republic of Korea |
| Abstract: |
Most of the current automatic speech recognition is performed on a remote server. However, the demand for speech recognition on personal devices is increasing, owing to the requirement of shorter recognition latency and increased privacy. End-to-end speech recognition that employs recurrent neural networks (RNNs) shows good accuracy, but the execution of conventional RNNs, such as the long short-term memory (LSTM) or gated recurrent unit (GRU), demands many memory accesses, thus hindering its real-time execution on smart-phones or embedded systems. To solve this problem, we built an end-to-end acoustic model (AM) using linear recurrent units instead of LSTM or GRU and employed a multi-step parallel approach for reducing the number of DRAM accesses. The AM is trained with the connectionist temporal classification (CTC) loss, and the decoding is conducted using weighted finite-state transducers (WFSTs). The proposed system achieves x4.8 real-time speed when executed on a single core of an ARM CPU-based system. |