acoustic modelling
Recently Published Documents


TOTAL DOCUMENTS

179
(FIVE YEARS 42)

H-INDEX

17
(FIVE YEARS 3)

Author(s):  
Rajeev Rajan ◽  
Jisna Johnson ◽  
Noumida Abdul Kareem

2022 ◽  
Vol 105 ◽  
pp. 136-149
Author(s):  
Elwin van 't Wout ◽  
Seyyed R. Haqshenas ◽  
Pierre Gélat ◽  
Timo Betcke ◽  
Nader Saffari

Author(s):  
Alaa Ehab Sakran ◽  
Mohsen Rashwan ◽  
Sherif Mahdy Abdou

In this paper, automatic segmentation system was built using the Kaldi toolkit at phoneme level for Quran verses data set with a total speech corpus of (80 hours) and its corresponding text corpus respectively, with a size of 1100 recorded Quran verses of 100 non-Arab reciters. Initiated with the extraction of Mel Frequency Cepstral Coefficients MFCCs, the proceedings of the building of Language Model LM and Acoustic Model AM training phase continued until the Deep Neural Network DNN level by selecting 770 waves (70 reciters). The testing of the system was done using 220 waves (20 reciters), and concluded with the selection of the development data set which was 280 waves (10 reciters). Comparison was implemented between automatic and manual segmentation, and the results obtained for the test set was 99% and for the Development set was 99% with Time Delay Neural Networks TDNN based acoustic modelling.


Author(s):  
S. Parameshwara ◽  
N.B. Pradeep ◽  
B.K. Venkatesha ◽  
K.P. Prashanth

2021 ◽  
pp. 104915
Author(s):  
Shuangxia Shi ◽  
Jingyu Wang ◽  
Kongchao Liu ◽  
Guoyong Jin ◽  
Bin Xiao

Author(s):  
Lucile Gelin ◽  
Morgane Daniel ◽  
Julien Pinquier ◽  
Thomas Pellegrini

2021 ◽  
Author(s):  
Erfan Loweimi ◽  
Zoran Cvetkovic ◽  
Peter Bell ◽  
Steve Renals
Keyword(s):  

2021 ◽  
Vol 263 (4) ◽  
pp. 2665-2673
Author(s):  
Thomas Judd ◽  
Stefan Weigand ◽  
Jochen Schaal

The analysis of noise and acoustics in indoor spaces is often performed with geometrical methods from the ray-tracing family, such as the sound particle method. In general, these offer an acceptable balance between physical accuracy and computational effort, but models with large numbers of objects and high levels of detail can lead to long waits for results. In this paper, we consider methods to assist with the efficient analysis of such situations in the context of the sound particle diffraction model. A modern open-plan office and a large cathedral are used as example projects. We look at space partitioning strategies, adaptive placement of receivers in the form of mesh noise maps, and graphics-card-style hardware acceleration techniques, along with iterative modelling methods. The role of geometrical detail in the context of uncertainties in the input data, such as absorption and scattering coefficients, is also studied. From this, we offer a range of recommendations regarding the level-of-detail in acoustic modelling, including consideration of issues such as seating, tables, and curved surfaces.


2021 ◽  
Vol 26 (2) ◽  
pp. 95-102
Author(s):  
David R. Bergman

A connection between acoustic rays in a moving inhomogeneous fluid medium and the null geodesic of a pseudo-Riemannian manifold provides a mechanism to derive several well-known results commonly used in acoustic ray theory. Among these include ray integrals for depth dependent sound speed and current profiles commonly used in ocean and aero acoustic modelling. In this new paradigm these are derived by application of a symmetry of the effective metric tensor known as isometry. In addition to deriving well-known results, the application of the full machinery of differential geometry offers a unified approach to modelling acoustic fields in three dimensional random environments with time dependence by, (1) using conformal symmetry to simplify the geodesic equation, and (2) application of geodesic deviation as a generalization of geometric spread.


Sign in / Sign up

Export Citation Format

Share Document