%Aigaion2 BibTeX export from Idiap Publications
%Saturday 04 May 2024 05:45:18 AM

@INPROCEEDINGS{Nanchen_ICASSP_2019,
         author = {Nanchen, Alexandre and Garner, Philip N.},
       projects = {Idiap, SUMMA},
          month = may,
          title = {EMPIRICAL EVALUATION AND COMBINATION OF PUNCTUATION PREDICTION MODELS APPLIED TO BROADCAST NEWS},
      booktitle = {Proceedings of 2019 IEEE International Conference on Acoustics, Speech, and Signal Processing},
           year = {2019},
       crossref = {Nanchen_Idiap-RR-01-2019},
       abstract = {Natural language processing techniques are dependent upon punctuation to work well. When their input is taken from speech recognition, it is necessary to reconstruct the punctuation; in particular sentence boundaries. We define a range of features from low level acoustics to those with high level lexical semantics, including deep and recurrent models; these in turn are representative of a broad range of approaches used by previous authors for punctuation prediction. We combine the features using a gradient boosting machine that is also capable of indicating the relative importance of each feature. In an empirical study, we show that features from different semantic levels are in fact complementary, that combining statistical and deep learning methods yields better prediction results, and that generalization across different speaking styles is difficult to achieve without adaptation. Our best model achieves an F-Measure of 82.8 on a challenging broadcast news dataset.},
            pdf = {https://publications.idiap.ch/attachments/papers/2019/Nanchen_ICASSP_2019.pdf}
}



crossreferenced publications: 
@TECHREPORT{Nanchen_Idiap-RR-01-2019,
         author = {Nanchen, Alexandre and Garner, Philip N.},
       projects = {Idiap, SUMMA},
          month = {2},
          title = {EMPIRICAL EVALUATION AND COMBINATION OF PUNCTUATION PREDICTION MODELS APPLIED TO BROADCAST NEWS},
           type = {Idiap-RR},
         number = {Idiap-RR-01-2019},
           year = {2019},
    institution = {Idiap},
       abstract = {Natural language processing techniques are dependent upon punctuation to work well. When their input is taken from speech recognition, it is necessary to reconstruct the punctuation; in particular sentence boundaries. We define a range of features from low level acoustics to those with high level lexical semantics, including deep and recurrent models; these in turn are representative of a broad range of approaches used by previous authors for punctuation prediction. We combine the features using a gradient boosting machine that is also capable of indicating the relative importance of each feature. In an empirical study, we show that features from different semantic levels are in fact complementary, that combining statistical and deep learning methods yields better prediction results, and that generalization across different speaking styles is difficult to achieve without adaptation. Our best model achieves an F-Measure of 82.8 on a challenging broadcast news dataset.},
            pdf = {https://publications.idiap.ch/attachments/reports/2019/Nanchen_Idiap-RR-01-2019.pdf}
}