Cache based recurrent neural network language model inference for first pass speech recognition
ICASSP(2014)
摘要
Recurrent neural network language models (RNNLMs) have recently produced improvements on language processing tasks ranging from machine translation to word tagging and speech recognition. To date, however, the computational expense of RNNLMs has hampered their application to first pass decoding. In this paper, we show that by restricting the RNNLM calls to those words that receive a reasonable score according to a n-gram model, and by deploying a set of caches, we can reduce the cost of using an RNNLM in the first pass to that of using an additional n-gram model. We compare this scheme to lattice rescoring, and find that they produce comparable results for a Bing Voice search task. The best performance results from rescoring a lattice that is itself created with a RNNLM in the first pass.
更多查看译文
关键词
bing voice search task,lattice rescoring,recurrent neural network language model,speech recognition,language processing tasks,cache based recurrent neural network language model inference,n-gram model,computational efficiency,computational expense,rnnlm calls,language translation,cache,search problems,word tagging,recurrent neural nets,voice search,first pass speech recognition,machine translation,history,recurrent neural networks,computational modeling,decoding,data models,hidden markov models
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络