arXiv subjects

R. Coelho

Publications and source records attributed to R. Coelho.

16 recordsLinked to original sources

arXiv2024

Harmonic Detection from Noisy Speech with Auditory Frame Gain for Intelligibility Enhancement

This paper introduces a novel (HDAG - Harmonic Detection for Auditory Gain) method for speech intelligibility enhancement in noisy scenarios. In the proposed scheme, a series of selective Gammachirp filters are adopted to emphasize the harmonic components of speech reducing the masking effects of acoustic noises. The fundamental frequency are estimated by the HHT-Amp technique. Harmonic patterns estimated with low accuracy are detected and adjusted according the FSFFE low/high pitch separation. The central frequencies of the filterbank are defined considering the third octave subbands which are best suited to cover the regions most relevant to intelligibility. Before signal reconstruction, the gammachirp filtered components are amplified by gain factors regulated by FSFFE classification. The proposed HDAG solution and three baseline techniques are examined considering six background noises with four signal-to-noise ratios. Three objective measures are adopted for the evaluation of speech intelligibility and quality. Several experiments are conducted to demonstrate that the proposed scheme achieves better speech intelligibility improvement when compared to the competing approaches. A perceptual listening test is further considered and corroborates with the objective results.

eess.AS↗

arXiv2021

Noisy Speech Based Temporal Decomposition to Improve Fundamental Frequency Estimation

This paper introduces a novel method to separate noisy speech into low or high frequency frames, in order to improve fundamental frequency (F0) estimation accuracy. In this proposal, the target signal is analyzed by means of the ensemble empirical mode decomposition. Next, the pitch information is extracted from the first decomposition modes. This feature indicates the frequency region where the F0 of speech should be located, thus separating the frames into low-frequency (LF) or high-frequency (HF). The separation is applied to correct candidates extracted from a conventional fundamental frequency detection method, and hence improving the accuracy of F0 estimate. The proposed method is evaluated in experiments with CSTR and TIMIT databases, considering six acoustic noises under various signal-to-noise ratios. A pitch enhancement algorithm is adopted as baseline in the evaluation analysis considering three conventional estimators. Results show that the proposed method outperforms the competing strategies, in terms of low/high frequency separation accuracy. Moreover, the performance metrics of the F0 estimation techniques show that the novel solution is able to better improve F0 detection accuracy when compared to competitive approaches under different noisy conditions.

eess.AS↗

arXiv2021

Harmonic and non-Harmonic Based Noisy Reverberant Speech Enhancement in Time Domain

This paper introduces the single step time domain method named HnH-NRSE, whihc is designed for simultaneous speech intelligibility and quality improvement under noisy-reverberant conditions. In this solution, harmonic and non-harmonic elements of speech are separated by applying zero-crossing and energy criteria. An objective evaluation of the its non-stationarity degree is further used for an adaptive gain to treat masking components. No prior knowledge of speech statistics or room information is required for this technique. Additionally, two combined solutions, IRMO and IRMN, are proposed as composite methods for improvement on noisy-reverberant speech signals. The proposed and baseline methods are evaluated considering two intelligibility and three quality measures, applied for the objective prediction. The results show that the proposed scheme leads to a higher intelligibility and quality improvement when compared to competing methods in most scenarios. Additionally, a perceptual intelligibility listening test is performed, which corroborates with these results. Furthermore, the proposed HnH-NRSE solution attains SRMR quality measure with similar results when compared to the composed IRMO and IRMN techniques.

eess.AS↗

arXiv2021

New plasma regimes with small ELMs and high confinement at the Joint European Torus

New plasma regimes with high confinement, low core impurity accumulation and small Edge localized mode (ELMs) perturbations have been obtained close to ITER conditions in magnetically confined plasmas from the Joint European torus (JET) tokamak. Such regimes are achieved by means of optimized particle fuelling conditions which trigger a self-organize state with a strong increase in rotation and ion temperature and a decrease of the edge density. An interplay between core and edge plasma regions leads to reduced turbulence levels and outward impurity convection. These results pave the way to an attractive alternative to the standard plasmas considered for fusion energy generation in a tokamak with metallic wall environment such as the ones expected in ITER

physics.plasm-ph↗

arXiv2020

F0-based Gammatone Filtering for Intelligibility Gain of Acoustic Noisy Signals

This paper proposes a time-domain method to improve speech intelligibility in noisy scenarios. In the proposed approach, a series of Gammatone filters are adopted to detect the harmonic components of speech. The filters outputs are amplified to emphasize the first harmonics, reducing the masking effects of acoustic noises. The proposed GTFF0 solution and two baseline techniques are examined considering four background noises with different non-stationarity degrees. Three intelligibility measures (ESTOI, ESII and ASIIST) are adopted for objective evaluation. The experiments results show that the proposed scheme leads to expressive speech intelligibility gain when compared to the competing approaches. Furthermore, the PESQ and WSS objective scores demonstrate that the proposed technique also provides interesting quality improvement.

eess.AS↗

arXiv2020

Blind Mask to Improve Intelligibility of Non-Stationary Noisy Speech

This letter proposes a novel blind acoustic mask (BAM) designed to adaptively detect noise components and preserve target speech segments in time-domain. A robust standard deviation estimator is applied to the non-stationary noisy speech to identify noise masking elements. The main contribution of the proposed solution is the use of this noise statistics to derive an adaptive information to define and select samples with lower noise proportion. Thus, preserving speech intelligibility. Additionally, no information of the target speech and noise signals statistics is previously required to this non-ideal mask. The BAM and three competitive methods, Ideal Binary Mask (IBM), Target Binary Mask (TBM), and Non-stationary Noise Estimation for Speech Enhancement (NNESE), are evaluated considering speech signals corrupted by three non-stationary acoustic noises and six values of signal-to-noise ratio (SNR). Results demonstrate that the BAM technique achieves intelligibility gains comparable to ideal masks while maintaining good speech quality.

eess.AS↗

arXiv2019

Effective Acoustic Energy Sensing Exploitation for Target Sources Localization in Urban Acoustic Scenes

This letter proposes a new approach to improve the accuracy of the Energy-based source localization methods in urban acoustic scenes. The proposed acoustic energy sensing flow estimation (ESFE) uses the sensors signal nonstationarity degree to determine the area with highest energy concentration in the scenes. The ESFE is applied to different acoustic scenes and yields to source localization accuracy improvement with computational complexity reduction. The experiments results show that the proposed scheme leads to significant improvement in source localization accuracy.

eess.AS↗

arXiv2019

Impulsive Noise Detection for Intelligibility and Quality Improvement of Speech Enhancement Methods Applied in Time-Domain

This letter introduces a novel speech enhancement method in the Hilbert-Huang Transform domain to mitigate the effects of acoustic impulsive noises. The estimation and selection of noise components is based on the impulsiveness index of decomposition modes. Speech enhancement experiments are conducted considering five acoustic noises with different impulsiveness index and non-stationarity degrees under various signal-to-noise ratios. Three speech enhancement algorithms are adopted as baseline in the evaluation analysis considering spectral and time domains. The proposed solution achieves the best results in terms of objective quality measures and similar speech intelligibility rates to the competitive methods.

eess.AS↗

arXiv2019

Adaptive Reverberation Absorption using Non-stationary Masking Components Detection for Intelligibility Improvement

This letter proposes a new time domain absorption approach designed to reduce masking components of speech signals under noisy-reverberant conditions. In this method, the non-stationarity of corrupted signal segments is used to detect masking distortions based on a defined threshold. The nonstationarity is objectively measured and is also adopted to determine the absorption procedure. Additionally, no prior knowledge of speech statistics or of the room information is required for this technique. Three intelligibility measures (ESII, ASIIST, SRMRnorm) and a perceptual listening test are used for evaluation. The experiments results show that the proposed scheme leads to a higher intelligibility improvement when compared to competing methods.

eess.AS↗

arXiv2019

Objective Human Affective Vocal Expression Detection and Automatic Classification with Stochastic Models and Learning Systems

This paper presents a widespread analysis of affective vocal expression classification systems. In this study, state-of-the-art acoustic features are compared to two novel affective vocal prints for the detection of emotional states: the Hilbert-Huang-Hurst Coefficients (HHHC) and the vector of index of non-stationarity (INS). HHHC is here proposed as a nonlinear vocal source feature vector that represents the affective states according to their effects on the speech production mechanism. Emotional states are highlighted by the empirical mode decomposition (EMD) based method, which exploits the non-stationarity of the affective acoustic variations. Hurst coefficients (closely related to the excitation source) are then estimated from the decomposition process to compose the feature vector. Additionally, the INS vector is introduced as dynamic information to the HHHC feature. The proposed features are evaluated in speech emotion classification experiments with three databases in German and English languages. Three state-of-the-art acoustic features are adopted as baseline. The $α$-integrated Gaussian model ($α$-GMM) is also introduced for the emotion representation and classification. Its performance is compared to competing stochastic and machine learning classifiers. Results demonstrate that HHHC leads to significant classification improvement when compared to the baseline acoustic features. Moreover, results also show that $α$-GMM outperforms the competing classification methods. Finally, HHHC and INS are also evaluated as complementary features for the GeMAPS and eGeMAPS feature sets

eess.AS↗

arXiv2016

Sensitivity of alpha-particle-driven Alfven eigenmodes to q-profile variation in ITER scenarios

A perturbative hybrid ideal-MHD/drift-kinetic approach to assess the stability of alpha-particle-driven Alfven eigenmodes in burning plasmas is used to show that certain foreseen ITER scenarios, namely the Ip = 15 MA baseline scenario with very low and broad core magnetic shear, are sensitive to small changes in the background magnetic equilibrium. Slight variations (of the order of 1%) of the safety-factor value on axis are seen to cause large changes in the growth rate, toroidal mode number, and radial location of the most unstable eigenmodes found. The observed sensitivity is shown to proceed from the very low magnetic shear values attained throughout the plasma core, raising issues about reliable predictions of alpha-particle transport in burning plasmas.

physics.plasm-ph↗

arXiv2016

Comprehensive evaluation of the linear stability of Alfvén eigenmodes driven by alpha particles in an ITER baseline scenario

The linear stability of Alfvén eigenmodes in the presence of fusion-born alpha particles is thoroughly assessed for two variants of an ITER baseline scenario, which differ significantly in their core and pedestal temperatures. A systematic approach is used that considers all possible eigenmodes for a given magnetic equilibrium and determines their growth rates due to alpha-particle drive and Landau damping on fuel ions, helium ashes and electrons. This extensive stability study is efficiently conducted through the use of a specialized workflow that profits from the performance of the hybrid MHD drift-kinetic code $\mbox{CASTOR-K}$ (Borba D. and Kerner W. 1999 J. Comput. Phys. ${\bf 153}$ 101; Nabais F. ${\it et\,al}$ 2015 Plasma Sci. Technol. ${\bf 17}$ 89), which can rapidly evaluate the linear growth rate of an eigenmode. It is found that the fastest growing instabilities in the aforementioned ITER scenario are core-localized, low-shear toroidal Alfvén eigenmodes. The largest growth-rates occur in the scenario variant with higher core temperatures, which has the highest alpha-particle density and density gradient, for eigenmodes with toroidal mode numbers $n\approx30$. Although these eigenmodes suffer significant radiative damping, which is also evaluated, their growth rates remain larger than those of the most unstable eigenmodes found in the variant of the ITER baseline scenario with lower core temperatures, which have $n\approx15$ and are not affected by radiative damping.

physics.plasm-ph↗

arXiv2014

A systematic approach to the linear-stability assessment of Alfvén eigenmodes in the presence of fusion-born alpha particles for ITER-like equilibria

A systematic approach to assess the linear stability of Alfvén eigenmodes in the presence of fusion-born alpha particles is described. Because experimental results for ITER are not available yet, it is not known beforehand which Alfvén eigenmodes will interact more intensively with the alpha-particle population. Therefore, the number of modes that need to be considered in stability assessments becomes quite large and care must be exercised when choosing the numerical tools to work with, which must be fast and efficient. In the presented approach, all possible eigenmodes are first found after intensively scanning a suitable frequency range. Each solution found is then tested to find if its discretization over the radial grid in use is adequate. Finally, the interaction between the identified eigenmodes and the alpha-particle population is evaluated with the drift-kinetic code CASTOR-K, in order to assess their growth rates and hence their linear stability. The described approach enables one to single out the most unstable eigenmodes in a given scenario, which can then be handled with more specialized tools. This selection capability eases the task of evaluating alpha-particle interactions with Alfvén eigenmodes, either for ITER scenarios or for any other kind of scenario planning.

physics.plasm-ph↗

arXiv2014

Contrasting H-mode behaviour with deuterium fuelling and nitrogen seeding in the all-carbon and metallic versions of JET

The former all-carbon wall on JET has been replaced with beryllium in the main torus and tungsten in the divertor to mimic the surface materials envisaged for ITER. Comparisons are presented between Type I H-mode characteristics in each design by examining respective scans over deuterium fuelling and impurity seeding, required to ameliorate exhaust loads both in JET at full capability and in ITER.

physics.plasm-ph↗

arXiv2006

Sector analysis for a FTSE portfolio of stocks

Using a portfolio of stocks from the London Stock Exchange FTSE100 index (FTSE), we study both the time dependence of their correlations and the normalized tree length of the associated minimal spanning tree (MST). The first four moments of the distribution of correlations and lengths of the tree are examined in detail and differences in behaviour noted. For different economic groups and industries, clustering is evident. However comparing the classification used prior to 2006 with that introduced in January 2006 it is clear that the new classification, apart from one or two notable exceptions, is much more compatible with the clustering obtained by the MST analysis. We finally compare the MST for real data with that obtained for a synthetic {\it random market}. The latter tree would seem more like the structure found by Coronnello {\it et al.} for trees based on high frequency data.

physics.soc-ph↗

arXiv2004

A family-network model for wealth distribution in societies

A model based on first-degree family relations network is used to describe the wealth distribution in societies. The network structure is not a-priori introduced in the model, it is generated in parallel with the wealth values through simple and realistic dynamical rules. The model has two main parameters, governing the wealth exchange in the network. Choosing their values realistically, leads to wealth distributions in good agreement with measured data. The cumulative wealth distribution function has an exponential behavior in the low and medium wealth limit, and shows the Pareto-like power-law tail for the upper 5% of the society. The obtained Pareto indexes are in good agreement with the measured ones. The generated family networks also converges to a statistically stable topology with a simple Poissonian degree distribution. On this family-network many interesting correlations are studied, and the main factors leading to wealth-diversification and the formation of the Pareto law are identified.

cond-mat.other↗