| A Generative Adversarial Net-Based Bandwidth Extension Method for Audio Compression | |
Huang, Qingbo1,2; Liu TJ(刘铁军)2 ; Wu XH(吴玺宏)1; Qu TS(曲天书)1; Huang, Qingbo3,4; Liu TJ(刘铁军)4; Wu XH(吴玺宏)3; Qu TS(曲天书)3
| |
| Department | 水下机器人研究室 |
| Source Publication | JOURNAL OF THE AUDIO ENGINEERING SOCIETY
 |
| ISSN | 1549-4950 |
| 2019 | |
| Volume | 67Issue:12Pages:986-993 |
| Indexed By | SCI ; EI |
| EI Accession number | 20200508105153 |
| WOS ID | WOS:000505043700007 |
| Contribution Rank | 1 |
| Funding Organization | State Key Laboratory of Robotics [2018-O09] ; National Natural Science Foundation of ChinaNational Natural Science Foundation of China [61175043, 61421062] ; High Performance Computing Platform of Peking University |
| Abstract | The high frequency components of the audio signal are often truncated during the encoding processing by a lossy codec. To avoid the sound quality degradation, the high frequency components are reconstructed during the decoding processing. This paper presents a new bandwidth extension method for audio compression. Frequency components of 6.9 -13.8 kHz are added using side information at 2 kbps. A generative neural network in the GAN is used to estimate relationship between the MDCT spectrum in the high frequency part and the low frequency part, and it is evaluated by a discriminant network in the GAN to get a more natural result. On this basis, a codec system is built up. The MUSHRA experiments show that the proposed method is comparable with SBR in HE-AAC. |
| Language | 英语 |
| WOS Subject | Acoustics ; Engineering, Multidisciplinary |
| WOS Keyword | NARROW-BAND ; SPEECH |
| WOS Research Area | Acoustics ; Engineering |
| Funding Project | State Key Laboratory of Robotics[2018-O09] ; National Natural Science Foundation of China[61175043] ; National Natural Science Foundation of China[61421062] ; High Performance Computing Platform of Peking University |
| Citation statistics | |
| Document Type | 期刊论文 |
| Identifier | http://ir.sia.cn/handle/173321/26182 |
| Collection | 水下机器人研究室 |
| Corresponding Author | Qu TS(曲天书); Qu TS(曲天书) |
| Affiliation | 1.Peking Univ, Speech & Hearing Res Ctr, Minist Educ, Key Lab Machine Percept, Beijing, Peoples R China 2.Chinese Acad Sci, Shenyang Inst Automat, State Key Lab Robot, Beijing, Peoples R China 3.Peking Univ, Speech & Hearing Res Ctr, Minist Educ, Key Lab Machine Percept, Beijing, Peoples R China 4.Chinese Acad Sci, Shenyang Inst Automat, State Key Lab Robot, Beijing, Peoples R China |
| Recommended Citation GB/T 7714 | Huang, Qingbo,Liu TJ,Wu XH,et al. A Generative Adversarial Net-Based Bandwidth Extension Method for Audio Compression[J]. JOURNAL OF THE AUDIO ENGINEERING SOCIETY,2019,67(12):986-993. |
| APA | Huang, Qingbo.,Liu TJ.,Wu XH.,Qu TS.,Huang, Qingbo.,...&Qu TS.(2019).A Generative Adversarial Net-Based Bandwidth Extension Method for Audio Compression.JOURNAL OF THE AUDIO ENGINEERING SOCIETY,67(12),986-993. |
| MLA | Huang, Qingbo,et al."A Generative Adversarial Net-Based Bandwidth Extension Method for Audio Compression".JOURNAL OF THE AUDIO ENGINEERING SOCIETY 67.12(2019):986-993. |
| Files in This Item: | ||||||
| File Name/Size | DocType | Version | Access | License | ||
| A Generative Adversa（2209KB） | 期刊论文 | 作者接受稿 | 开放获取 | CC BY-NC-SA | View Application Full Text | |
Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.
Edit Comment