Acoustic DOA estimation using space alternating sparse Bayesian learning
| Autor: | Jesper Jensen, Mads Græsbøll Christensen, Liming Shi, Jinwei Sun, Zonglong Bai |
|---|---|
| Jazyk: | angličtina |
| Rok vydání: | 2021 |
| Předmět: |
Acoustics and Ultrasonics
Computational complexity theory Mean squared error Computer science Bayesian probability lcsh:QC221-246 02 engineering and technology Bayesian inference lcsh:QA75.5-76.95 Sound source localization 030507 speech-language pathology & audiology 03 medical and health sciences 0202 electrical engineering electronic engineering information engineering Electrical and Electronic Engineering Acoustic DOA estimation Scalar (physics) Estimator 020206 networking & telecommunications Hidden variable theory lcsh:Acoustics. Sound Sparse Bayesian learning lcsh:Electronic computers. Computer science 0305 other medical science Algorithm Humanoid robot |
| Zdroj: | Bai, Z, Shi, L, Jensen, J R, Sun, J & Christensen, M G 2021, ' Acoustic DOA estimation using space alternating sparse Bayesian learning ', Eurasip Journal on Audio, Speech, and Music Processing, vol. 2021, no. 1, 14, pp. 1-19 . https://doi.org/10.1186/s13636-021-00200-z EURASIP Journal on Audio, Speech, and Music Processing, Vol 2021, Iss 1, Pp 1-19 (2021) |
| Popis: | Estimating the direction-of-arrival (DOA) of multiple acoustic sources is one of the key technologies for humanoidrobots and drones. However, it is a most challenging problem due to a number of factors, including the platform sizewhich puts a constraint on the array aperture. To overcome this problem, a high-resolution DOA estimation algorithmbased on sparse Bayesian learning is proposed in this paper. A group sparse prior based hierarchical Bayesian model isintroduced to encourage spatial sparsity of acoustic sources. To obtain approximate posteriors of the hiddenvariables, a variational Bayesian approach is proposed. Moreover, to reduce the computational complexity, the spacealternating approach is applied to push the variational Bayesian inference to the scalar level. Furthermore, an acousticDOA estimator is proposed to jointly utilize the estimated source signals from all frequency bins. Compared tostate-of-the-art approaches, the high-resolution performance of the proposed approach is demonstrated inexperiments with both synthetic and real data. The experiments show that the proposed approach achieves lowerroot mean square error (RMSE), false alert (FA), and miss-detection (MD) than other methods. Therefore, the proposedapproach can be applied to some applications such as humanoid robots and drones to improve the resolutionperformance for acoustic DOA estimation especially when the size of the array aperture is constrained by theplatform, preventing the use of traditional methods to resolve multiple sources. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|