1 |
Challenges in Audio Processing of Terrorist-Related Data
|
|
|
|
In: International Conference on Multimedia Modeling ; https://hal.archives-ouvertes.fr/hal-02415176 ; International Conference on Multimedia Modeling, Springer, Jan 2019, Thessaloniki, Greece (2019)
|
|
BASE
|
|
Show details
|
|
2 |
Challenges in Audio Processing of Terrorist-Related Data
|
|
|
|
In: International Conference on Multimedia Modeling ; https://hal.archives-ouvertes.fr/hal-02387373 ; International Conference on Multimedia Modeling, Springer, Jan 2019, Thessaloniki, Greece (2019)
|
|
BASE
|
|
Show details
|
|
3 |
An investigation into language model data augmentation for low-resourced STT and KWS
|
|
|
|
In: IEEE International Conference on Acoustics, Speech, and Signal Processing ; https://hal.archives-ouvertes.fr/hal-01837171 ; IEEE International Conference on Acoustics, Speech, and Signal Processing, IEEE, Mar 2017, New Orleans, United States (2017)
|
|
Abstract:
International audience ; This paper reports on investigations using two techniques for language model text dataaugmentation for low-resourced automatic speech recognition and keyword search. Low-resourced languages are characterized by limited training materials, which typically resultsin high out-of-vocabulary (OOV) rates and poor language model estimates. One techniquemakes use of recurrent neural networks (RNNs) using word or subword units. Word-basedRNNs keep the same system vocabulary, so they cannot reduce the OOV, whereas subwordunits can reduce the OOV but generate many false combinations. A complementarytechnique is based on automatic machine translation, which requires parallel texts and isable to add words to the vocabulary. These methods were assessed on 10 languages in thecontext of the Babel program and NIST OpenKWS evaluation. Although improvements vary across languages with both methods, small gains were generally observed in terms of word error rate reduction and improved keyword search performance.
|
|
Keyword:
[INFO.INFO-CL]Computer Science [cs]/Computation and Language [cs.CL]; [INFO]Computer Science [cs]; keyword search; low resourced languages; multilingual; speech recognition
|
|
URL: https://hal.archives-ouvertes.fr/hal-01837171
|
|
BASE
|
|
Hide details
|
|
4 |
Comparing decoding strategies for subword-based keyword spotting in low-resourced languages
|
|
|
|
In: Annual Conference of the International Speech Communication Association ; https://hal.archives-ouvertes.fr/hal-01843408 ; Annual Conference of the International Speech Communication Association , ISCA, Sep 2014, Singapore, Singapore (2014)
|
|
BASE
|
|
Show details
|
|
|
|