No abstract

No abstract

No abstract

Recognition of Hungarian conversational telephone speech is challenging due to the informal style and morphological richness of the language. Recurrent Neural Network Language Model (RNNLM) can provide remedy for the high perplexity of the task; however, twopass decoding introduces a considerable processing delay. In order to eliminate this delay we investigate approaches aiming at the complexity reduction of RNNLM, while preserving its accuracy. We compare the performance of conventional back-off n-gram language models (BNLM), BNLM approximation of RNNLMs (RNN-BNLM) and RNN n-grams in terms of perplexity and word error rate (WER). Morphological richness is often addressed by using statistically derived subwords -morphs -in the language models, hence our investigations are extended to morph-based models, as well. We found that using RNN-BNLMs 40% of the RNNLM perplexity reduction can be recovered, which is roughly equal to the performance of a RNN 4-gram model. Combining morph-based modeling and approximation of RNNLM, we were able to achieve 8% relative WER reduction and preserve real-time operation of our conversational telephone speech recognition system.

Freely available audiobooks are a rich resource of expressive speech recordings that can be used for the purposes of speech synthesis. Natural sounding, expressive synthetic voices have previously been built from audiobooks that contained large amounts of highly expressive speech recorded from a profes- sionally trained speaker. The majority of freely available au- diobooks, however, are read by amateur speakers, are shorter and contain less expressive (less emphatic, less emotional, etc.) speech both in terms of quality and quantity. Synthesiz- ing expressive speech from a typical online audiobook there- fore poses many challenges. In this work we address these challenges by applying a method consisting of minimally su- pervised techniques to align the text with the recorded speech, select groups of expressive speech segments and build expres- sive voices for hidden Markov-model based synthesis using speaker adaptation. Subjective listening tests have shown that the expressive synthetic speech generated with this method is often able to produce utterances suited to an emotional mes- sage. We used a restricted amount of speech data in our exper- iment, in order to show that the method is generally applicable to most typical audiobooks widely available online. 

QC 20160422