Name
Title | ||
|---|---|---|
Multitask Learning With Low-Level Auxiliary Tasks For Encoder-Decoder Based Speech Recognition |
Abstract | ||
|---|---|---|
End-to-end training of deep learning-based models allows for implicit learning of intermediate representations based on the final task loss. However, the end-to-end approach ignores the useful domain knowledge encoded in explicit intermediate-level supervision. We hypothesize that using intermediate representations as auxiliary supervision at lower levels of deep networks may be a good way of combining the advantages of end-to-end training and more traditional pipeline approaches. We present experiments on conversational speech recognition where we use lower-level tasks, such as phoneme recognition, in a multitask training approach with an encoder-decoder model for direct character transcription. We compare multiple types of lower-level tasks and analyze the effects of the auxiliary tasks. Our results on the Switchboard corpus show that this approach improves recognition accuracy over a standard encoder-decoder model on the Eva12000 test set. |
Year | DOI | Venue |
|---|---|---|
2017 | 10.21437/Interspeech.2017-1118 | 18TH ANNUAL CONFERENCE OF THE INTERNATIONAL SPEECH COMMUNICATION ASSOCIATION (INTERSPEECH 2017), VOLS 1-6: SITUATED INTERACTION |
Keywords | DocType | Volume |
speech recognition, multitask learning, encoder-decoder, CTC, LSTM | Conference | abs/1704.01631 |
ISSN | Citations | PageRank |
2308-457X | 10 | 0.54 |
References | Authors | |
16 | 4 | |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Shubham Toshniwal | 1 | 19 | 4.12 |
| Hao Tang | 2 | 43 | 5.30 |
| Liang Lu | 3 | 894 | 165.81 |
| Karen Livescu | 4 | 1254 | 71.43 |