Formant Tracking Using Quasi-Closed Phase Forward-Backward Linear Prediction Analysis and Deep Neural Networks
Formant tracking is investigated in this study by using trackers based on dynamic programming (DP) and deep neural nets (DNNs). Using the DP approach, six formant estimation methods were first compared. The six methods include linear prediction (LP) algorithms, weighted LP algorithms and the recentl...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2021-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/9606741/ |
| _version_ | 1818822741815459840 |
|---|---|
| author | Dhananjaya N. Gowda Bajibabu Bollepalli Sudarsana Reddy Kadiri Paavo Alku |
| author_facet | Dhananjaya N. Gowda Bajibabu Bollepalli Sudarsana Reddy Kadiri Paavo Alku |
| author_sort | Dhananjaya N. Gowda |
| collection | DOAJ |
| description | Formant tracking is investigated in this study by using trackers based on dynamic programming (DP) and deep neural nets (DNNs). Using the DP approach, six formant estimation methods were first compared. The six methods include linear prediction (LP) algorithms, weighted LP algorithms and the recently developed quasi-closed phase forward-backward (QCP-FB) method. QCP-FB gave the best performance in the comparison. Therefore, a novel formant tracking approach, which combines benefits of deep learning and signal processing based on QCP-FB, was proposed. In this approach, the formants predicted by a DNN-based tracker from a speech frame are refined using the peaks of the all-pole spectrum computed by QCP-FB from the same frame. Results show that the proposed DNN-based tracker performed better both in detection rate and estimation error for the lowest three formants compared to reference formant trackers. Compared to the popular Wavesurfer, for example, the proposed tracker gave a reduction of 29%, 48%, and 35% in the estimation error for the lowest three formants, respectively. |
| first_indexed | 2024-12-18T23:28:54Z |
| format | Article |
| id | doaj.art-d3e5f843aaf44751a9b53d75a6163f7e |
| institution | Directory Open Access Journal |
| issn | 2169-3536 |
| language | English |
| last_indexed | 2024-12-18T23:28:54Z |
| publishDate | 2021-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj.art-d3e5f843aaf44751a9b53d75a6163f7e2022-12-21T20:47:44ZengIEEEIEEE Access2169-35362021-01-01915163115164010.1109/ACCESS.2021.31262809606741Formant Tracking Using Quasi-Closed Phase Forward-Backward Linear Prediction Analysis and Deep Neural NetworksDhananjaya N. Gowda0Bajibabu Bollepalli1Sudarsana Reddy Kadiri2https://orcid.org/0000-0001-5806-3053Paavo Alku3https://orcid.org/0000-0002-8173-9418Samsung Research, Seoul, South KoreaAmazon, Slough, U.K.Department of Signal Processing and Acoustics, Aalto University, Espoo, FinlandDepartment of Signal Processing and Acoustics, Aalto University, Espoo, FinlandFormant tracking is investigated in this study by using trackers based on dynamic programming (DP) and deep neural nets (DNNs). Using the DP approach, six formant estimation methods were first compared. The six methods include linear prediction (LP) algorithms, weighted LP algorithms and the recently developed quasi-closed phase forward-backward (QCP-FB) method. QCP-FB gave the best performance in the comparison. Therefore, a novel formant tracking approach, which combines benefits of deep learning and signal processing based on QCP-FB, was proposed. In this approach, the formants predicted by a DNN-based tracker from a speech frame are refined using the peaks of the all-pole spectrum computed by QCP-FB from the same frame. Results show that the proposed DNN-based tracker performed better both in detection rate and estimation error for the lowest three formants compared to reference formant trackers. Compared to the popular Wavesurfer, for example, the proposed tracker gave a reduction of 29%, 48%, and 35% in the estimation error for the lowest three formants, respectively.https://ieeexplore.ieee.org/document/9606741/Speech analysisformant trackinglinear predictiondynamic programmingdeep neural net |
| spellingShingle | Dhananjaya N. Gowda Bajibabu Bollepalli Sudarsana Reddy Kadiri Paavo Alku Formant Tracking Using Quasi-Closed Phase Forward-Backward Linear Prediction Analysis and Deep Neural Networks IEEE Access Speech analysis formant tracking linear prediction dynamic programming deep neural net |
| title | Formant Tracking Using Quasi-Closed Phase Forward-Backward Linear Prediction Analysis and Deep Neural Networks |
| title_full | Formant Tracking Using Quasi-Closed Phase Forward-Backward Linear Prediction Analysis and Deep Neural Networks |
| title_fullStr | Formant Tracking Using Quasi-Closed Phase Forward-Backward Linear Prediction Analysis and Deep Neural Networks |
| title_full_unstemmed | Formant Tracking Using Quasi-Closed Phase Forward-Backward Linear Prediction Analysis and Deep Neural Networks |
| title_short | Formant Tracking Using Quasi-Closed Phase Forward-Backward Linear Prediction Analysis and Deep Neural Networks |
| title_sort | formant tracking using quasi closed phase forward backward linear prediction analysis and deep neural networks |
| topic | Speech analysis formant tracking linear prediction dynamic programming deep neural net |
| url | https://ieeexplore.ieee.org/document/9606741/ |
| work_keys_str_mv | AT dhananjayangowda formanttrackingusingquasiclosedphaseforwardbackwardlinearpredictionanalysisanddeepneuralnetworks AT bajibabubollepalli formanttrackingusingquasiclosedphaseforwardbackwardlinearpredictionanalysisanddeepneuralnetworks AT sudarsanareddykadiri formanttrackingusingquasiclosedphaseforwardbackwardlinearpredictionanalysisanddeepneuralnetworks AT paavoalku formanttrackingusingquasiclosedphaseforwardbackwardlinearpredictionanalysisanddeepneuralnetworks |