Filters








2 Hits in 3.1 sec

Word Error Rate Estimation Without ASR Output: e-WER2 [article]

Ahmed Ali, Steve Renals
2020 arXiv   pre-print
Measuring the performance of automatic speech recognition (ASR) systems requires manually transcribed data in order to compute the word error rate (WER), which is often time-consuming and expensive.  ...  The estimated overall WER by e-WER2 is 30.9% for a three hours test set, while the WER computed using the reference transcriptions was 28.5%.  ...  Our novel system is able to learn acoustic-lexical embeddings to estimate the error rate directly without having access to the ASR results nor the ASR system -this is our "nobox" WER estimation method,  ... 
arXiv:2008.03403v1 fatcat:x572wpfb4fa6rpctjdm6qzkqnm

Word Error Rate Estimation Without ASR Output: e-WER2

Ahmed Ali, Steve Renals
2020 Interspeech 2020  
Measuring the performance of automatic speech recognition (ASR) systems requires manually transcribed data in order to compute the word error rate (WER), which is often timeconsuming and expensive.  ...  The estimated overall WER by e-WER2 is 30.9% for a three hours test set, while the WER computed using the reference transcriptions was 28.5%.  ...  Our novel system is able to learn acoustic-lexical embeddings to estimate the error rate directly without having access to the ASR results nor the ASR system -this is our "nobox" WER estimation method,  ... 
doi:10.21437/interspeech.2020-2357 dblp:conf/interspeech/0002R20 fatcat:7tg4ogbvsbcsbphlwdzbb37szy