Preliminary Experimental Study on Deep Neural Network-Based Dereverberation

Park JH, Jeon KM, Chun C, Yoo JS, Kim HK. Preliminary Experimental Study on Deep Neural Network-Based Dereverberation. In: AES Convention 141. Audio Engineering Society; 2016. Paper 300. Available from: https://aes.org/publications/elibrary-page/?id=18396

@inproceedings{Park2016_18396,
  author = {Park, Ji Hyun and Jeon, Kwang Myung and Chun, Chanjun and Yoo, Ji Sang and Kim, Hong Kook},
  title = {{Preliminary Experimental Study on Deep Neural Network-Based Dereverberation}},
  booktitle = {AES Convention 141},
  note = {Paper 300},
  year = {2016},
  month = sep,
  publisher = {Audio Engineering Society},
  url = {https://aes.org/publications/elibrary-page/?id=18396}
}

TY  - CPAPER
TI  - Preliminary Experimental Study on Deep Neural Network-Based Dereverberation
AU  - Park, Ji Hyun
AU  - Jeon, Kwang Myung
AU  - Chun, Chanjun
AU  - Yoo, Ji Sang
AU  - Kim, Hong Kook
T2  - AES Convention 141
M1  - Paper 300
PY  - 2016
DA  - 2016/09/06
UR  - https://aes.org/publications/elibrary-page/?id=18396
PB  - Audio Engineering Society
LA  - en
AB  - This paper deals with the issues associated with the dereverberation of speech or audio signals using deep neural networks (DNNs). They include feature extraction for DNNs from both clean and reverberant signals and DNN construction for generating dereverberant signals. To evaluate the performance of the proposed dereverberation method, artificially processed reverberant speech signals are obtained and a feed-forward DNN is constructed. It is shown that log spectral distortion (LSD) after applying DNN-based dereverberation is reduced by around 1.9 dB, compared with that of reverberant speech signals.
ER  -