Abstract | ||
---|---|---|
In this paper we introduce telephone noise removal methods used while we were developing a speech recognition system for formal sentence based paging system. It is a crucial factor to find the robust and simple computation methods for the actual application of telephone speech recognition. Channel compensation is an effective approach for robust telephone speech recognition. In this paper, we compare the general channel compensation method: CMS, two level CMS and RASTA for robust telephone speech recognition, and we also compare the performances of these former methods with that of our proposed method RMFCC. It is shown from the experiments that the proposed method, comparing with the RASTA processing, reduces the computational complexity without losing performance, and it is also better than CMS and two level CMS on the performance. After discussion, we also verify that it is an effective approach to suppress very low modulation frequencies for robust telephone speech recognition. |
Year | Venue | Field |
---|---|---|
1997 | PROGRESS IN CONNECTIONIST-BASED INFORMATION SYSTEMS, VOLS 1 AND 2 | Telephone network,Computer science,Speech recognition |
DocType | Citations | PageRank |
Conference | 0 | 0.34 |
References | Authors | |
0 | 4 |
Name | Order | Citations | PageRank |
---|---|---|---|
M. Han | 1 | 0 | 0.68 |
Gwi-Tae Park | 2 | 461 | 51.76 |
Park, J. | 3 | 2 | 1.78 |
J. Han | 4 | 0 | 0.34 |