A discriminative learning approach to probabilistic acoustic source localization

Kayser, Hendrik; Anemüller, Jörn

doi:10.1109/iwaenc.2014.6953346

Cited by 36 publications

(17 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To overcome the limitations of physical modeling, data-driven approaches propose to use supervised learning in order to grasp the complexity of acoustic phenomena. Pioneer works made use of kernel estimators [12], ridge regression [13], support vector machines [14] or Gaussian mixture models [15].…”

Section: Introductionmentioning

confidence: 99%

Regression Versus Classification for Neural Network Based Audio Source Localization

Perotin¹,

Défossez²,

Serizel³

et al. 2019

2019 IEEE Workshop on Applications of Signal Processing to Audio and Acoustics (WASPAA)

View full text Add to dashboard Cite

We compare the performance of regression and classification neural networks for single-source direction-of-arrival estimation. Since the output space is continuous and structured, regression seems more appropriate. However, classification on a discrete spherical grid is widely believed to perform better and is predominantly used in the literature. For regression, we propose two ways to account for the spherical geometry of the output space based either on the angular distance between spherical coordinates or on the mean squared error between Cartesian coordinates. For classification, we propose two alternatives to the classical one-hot encoding framework: we derive a Gibbs distribution from the squared angular distance between grid points and use the corresponding probabilities either as soft targets or as cross-entropy weights that retain a clear probabilistic interpretation. We show that regression on Cartesian coordinates is generally more accurate, except when localized interference is present, in which case classification appears to be more robust.

show abstract

Section: Introductionmentioning

confidence: 99%

Regression Versus Classification for Neural Network Based Audio Source Localization

Perotin¹,

Défossez²,

Serizel³

et al. 2019

2019 IEEE Workshop on Applications of Signal Processing to Audio and Acoustics (WASPAA)

View full text Add to dashboard Cite

show abstract

“…When combined with DOA estimation (e.g. [17,18]) that is not speech-specific, but instead produces estimates for any localized sound source in the acoustic scene, the entropy estimate could be used to differentiate between speech and non-speech sources, which might result in a benefit compared to methods that are not directly tailored to speech. In order to investigate more natural scenarios for hearing-impaired listeners, including additional data from accelerometers can be employed to compensate beam angles during head movements.…”

Section: Discussionmentioning

confidence: 99%

Assessing Speech Quality in Speech-Aware Hearing Aids Based on Phoneme Posteriorgrams

et al. 2016

Self Cite

View full text Add to dashboard Cite

Current behind-the-ear hearing aids (HA) allow to perform spatial filtering to enhance localized sound sources; however, they often lack processing strategies that are tailored to spoken language. Hence, without a feedback about speech quality achieved by the system, spatial filtering potentially remains unused, in case of a conservative enhancement strategy, or can even be detrimental to the speech intelligibility of the output signal. In this paper we apply phoneme posteriorgrams obtained from HA signals processed with deep neural networks to measure the quality of speech representations in spatial scenes. Inverse entropy of phoneme probabilities is proposed as a measure that allows to evaluate if current hearing aid parameters are optimal for the given acoustic condition. We investigate how varying noise levels and wrong estimates of the to-beenhanced direction affect this measure in anechoic and reverberant conditions and show our measure to provide a high reliability when varying each parameter.Experiments show that entropy as a function of the beam angle has a distinct minimum at the speaker's true position and its immediate vicinity. Thus, it can be used to determine the beam angle which optimizes the speech representation. Further, variations of the SNR cause a consistent offset of the entropy.

show abstract

“…The method employed here is a discriminative classification approach to probabilistic sound source localization [8]. It delivers the probability of the sound incidence for a defined set of source locations using short-term generalized cross-correlation functions [11] with phase transform (GCC-PHAT) as input features.…”

Section: Probabilistic Source Localizationmentioning

confidence: 99%

“…1 for an overview, consists of a signal enhancement front-end that uses spatial information about the target source which is obtained from interchannel phase differences weighted with spatial source probability. Spatial probability is estimated using a discriminative classification approach to source localization [8], which has been shown to be robust against noise and mismatch between room conditions in test and training data [9]. This spatial information is used as steering vector in an adaptive delay-andsum (DS) beamformer or in combination with noise statistics in a minimum-variance-distortionless-response (MVDR) beamformer.…”

Section: Introductionmentioning

confidence: 99%

Probabilistic Spatial Filter Estimation for Signal Enhancement in Multi-Channel Automatic Speech Recognition

2016

Self Cite

View full text Add to dashboard Cite

Speech recognition in multi-channel environments requires target speaker localization, multi-channel signal enhancement and robust speech recognition. We here propose a system that addresses these problems: Localization is performed with a recently introduced probabilistic localization method that is based on support-vector machine learning of GCC-PHAT weights and that estimates a spatial source probability map. The main contribution of the present work is the introduction of a probabilistic approach to (re-)estimation of location-specific steering vectors based on weighting of observed inter-channel phase differences with the spatial source probability map derived in the localization step. Subsequent speech recognition is carried out with a DNN-HMM system using amplitude modulation filter bank (AMFB) acoustic features which are robust to spectral distortions introduced during spatial filtering. The system has been evaluated on the CHIME-3 multichannel ASR dataset. Recognition was carried out with and without probabilistic steering vector re-estimation and with MVDR and delay-and-sum beamforming, respectively. Results indicate that the system attains on real-world evaluation data a relative improvement of 31.98% over the baseline and of 21.44% over a modified baseline. We note that this improvement is achieved without exploiting oracle knowledge about speech/non-speech intervals for noise covariance estimation (which is, however, assumed for baseline processing).

show abstract

A discriminative learning approach to probabilistic acoustic source localization

Cited by 36 publications

References 19 publications

Regression Versus Classification for Neural Network Based Audio Source Localization

Regression Versus Classification for Neural Network Based Audio Source Localization

Assessing Speech Quality in Speech-Aware Hearing Aids Based on Phoneme Posteriorgrams

Probabilistic Spatial Filter Estimation for Signal Enhancement in Multi-Channel Automatic Speech Recognition

Contact Info

Product

Resources

About