Abstract | ||
---|---|---|
This paper presents a simple end-to-end model for speech recognition, combining a convolutional network based acoustic model and a graph decoding. It is trained to output letters, with transcribed speech, without the need for force alignment of phonemes. We introduce an automatic segmentation criterion for training from sequence annotation without alignment that is on par with CTC (Graves et al., 2006) while being simpler. We show competitive results in word error rate on the Librispeech corpus (Panayotov et al., 2015) with MFCC features, and promising results from raw waveform. |
Year | Venue | Field |
---|---|---|
2016 | arXiv: Learning | Mel-frequency cepstrum,Computer science,End-to-end principle,Structured prediction,Artificial intelligence,Deep learning,Pattern recognition,Segmentation,Word error rate,Speech recognition,Decoding methods,Machine learning,Acoustic model |
DocType | Volume | Citations |
Journal | abs/1609.03193 | 32 |
PageRank | References | Authors |
1.09 | 13 | 3 |
Name | Order | Citations | PageRank |
---|---|---|---|
Ronan Collobert | 1 | 4002 | 308.61 |
Christian Puhrsch | 2 | 32 | 1.09 |
Gabriel Synnaeve | 3 | 240 | 16.91 |