Description of the UEDIN system for German ASR
J Driesen, P Bell, M Sinclair… - Proceedings of the 10th …, 2013 - research.ed.ac.uk
Proceedings of the 10th International Workshop on Spoken Language …, 2013•research.ed.ac.uk
In this paper we describe the ASR system for German built at the University of Edinburgh
(UEDIN) for the 2013 IWSLT evaluation campaign. For ASR, the major challenge to
overcome, was to find suitable acoustic training data. Due to the lack of expertly transcribed
acoustic speech data for German, acoustic model training had to be performed on publicly
available data crawled from the internet. For evaluation, lack of a manual segmentation into
utterances was handled in two different ways: by generating an automatic segmentation …
(UEDIN) for the 2013 IWSLT evaluation campaign. For ASR, the major challenge to
overcome, was to find suitable acoustic training data. Due to the lack of expertly transcribed
acoustic speech data for German, acoustic model training had to be performed on publicly
available data crawled from the internet. For evaluation, lack of a manual segmentation into
utterances was handled in two different ways: by generating an automatic segmentation …
Abstract
In this paper we describe the ASR system for German built at the University of Edinburgh (UEDIN) for the 2013 IWSLT evaluation campaign. For ASR, the major challenge to overcome, was to find suitable acoustic training data. Due to the lack of expertly transcribed acoustic speech data for German, acoustic model training had to be performed on publicly available data crawled from the internet. For evaluation, lack of a manual segmentation into utterances was handled in two different ways: by generating an automatic segmentation, and by treating entire input files as a single segment. Demonstrating the latter method is superior in the current task, we obtained a WER of 28.16% on the dev set and 36.21% on the test set.
research.ed.ac.uk
以上显示的是最相近的搜索结果。 查看全部搜索结果