2019_programme: SOURCE AZIMUTH ESTIMATION USING NEURAL NETWORK BASED ON SINGLE VECTOR SENOR: PRELIMINARY RESULTS

Back


  • Session: 22. Vector Hydrophone Research
    Organiser(s): Yang Desen
  • Lecture: SOURCE AZIMUTH ESTIMATION USING NEURAL NETWORK BASED ON SINGLE VECTOR SENOR: PRELIMINARY RESULTS [invited]
    Paper ID: 743
    Author(s): Cao Huaigang, Wang Wenbo, Ni Haiyan , Ma Li, Ren Qunyan
    Presenter: Ren Qunyan
    Presentation type: oral
    Abstract: Deep learning has been successfully used to estimate source depth and distance through two-step procedure of training and prediction, and here is adapted to estimate the source azimuth by analyzing ship noise data recorded by a single vector sensor. For the experiment, the vector was moored on the seafloor and the ship circled around the vector sensor. Source azimuth estimation can also be achieved by cross-spectral processing of the sound fields recorded on a single vector hydrophone, which exploits the physical properties of acoustic pressure and particle velocity components. Preliminary results demonstrated that the estimated source azimuth from these two approaches are in general agreement, however, their results are with poor resolution. As an effort to increase the accuracy of source azimuth estimation, the possibility of combing the physical properties of sound fields and deep learning algorithm is being studied.
  • Corresponding author: Dr Ren Qunyan
    Affiliation: Institute of Acoustics, Chinese Academy of Sciences
    Country: China
    e-mail:

Back