| Paper ID | F-2-3.7 |
| Paper Title |
Experimental investigation of robustness of spatial cepstrum features under various recording conditions |
| Authors |
Taiga Kawamura, Ryoichi Miyazaki, National Institute of Technology, Tokuyama College, Japan; Keisuke Imoto, Doshisha University, Japan; Nobutaka Ono, Tokyo Metropolitan University, Japan |
| Session |
F-2-3: Speech Enhancement 2 |
| Time | Wednesday, 09 December, 17:15 - 19:15 |
| Presentation Time: | Wednesday, 09 December, 18:45 - 19:00 Check your Time Zone |
|
All times are in New Zealand Time (UTC +13) |
| Topic |
Speech, Language, and Audio (SLA): |
| Abstract |
In recent years, devices that can easily record sounds, such as smartphones and tablets, have become widespread. Also, speech enhancement and acoustic scene analysis have been researched using a distributed microphone array consisting of these devices. The use of the spatial cepstrum has been proposed as a method for acquiring spatial information from a distributed microphone array. However, the behavior of the spatial cepstrum has been shown in the minimal experiment where the sound source is only white Gaussian noise, the microphone position is not changed, and the experiment is conducted under a specific reverberation condition. Therefore, we experimentally investigate the robustness of the spatial cepstrum under conditions closer to the real environment. The experimental results show that changes in the experimental conditions only affect the high-dimensional spatial cepstrum; thus, the low-dimensional spatial cepstrum is a robust feature that is not easily affected by disturbances. |