Timbre saliency vs. timbre dissimilarity ? What is the relationship?

Timbre spaces represent the organization of perceptual distances, as measured with dissimilarity ratings, among tones equated for pitch, loudness, and perceived duration. A number of potential acoustic correlates of timbre-space dimensions have been proposed in the psychoacoustic literature, including attack time, spectral centroid, spectral flux, and spectrum fine structure. The experiments reported here were designed as direct tests of the perceptual relevance of these acoustical parameters for timbre dissimilarity judgments. Listeners presented with carefully controlled synthetic tones use attack time, spectral centroid, and spectrum fine structure in dissimilarity rating experiments. These parameters thus appear as major determinants of timbre. However, spectral flux appears as a less salient timbre parameter, its salience depending on the number of other dimensions varying concurrently in the stimulus set. Dissimilarity ratings were analyzed with two different multidimensional scaling models (CLASCAL and CONSCAL), the latter providing psychophysical functions constrained by the physical parameters. Their complementarity is discussed.

Various methods have been used in an attempt to reveal salient parameters for musical timbre perception and to estimate perceptual distances between timbres based on their spectra. The multidimensional scaling method compares exemplars of musical sounds using listener estimation of timbral dissimilarity between sound pairs. Data are coalesced into a space whose dimensions are easily interpretable in terms of acoustic parameters such as spectral centroid, spectral flux, spectral irregularity, and attack time. Sounds can be normalized with respect to some parameters in order to reveal others. The direct comparison method attempts to discover how discrimination, recognition, and degree of perceptual dissimilarity depend on the degree of modification of the time-varying spectrum or of particular features of the spectrum. Morphing between spectrum pairs is one method of producing spectral modifications. Dissimilarity can be predicted by computing a “distance” between two time-varying spectra. For musical sounds, spectral distance has been computed by taking the average Cartesian distance between spectral measures treated as vectors. For constant-F0 sounds, the most obvious vector is the set of harmonic amplitudes, either in linear or decibel units. Other vectors are based on critical bands and cepstral coefficients. [Work supported by Research Grants Council Grant No. 613508.]

Experienced Listeners’ Perception of Timbre Dissimilarity Within and Between Voice Categories

Ten sustained musical instrument tones (bassoon, cello, clarinet, flute, horn, oboe, recorder, alto saxophone, trumpet, and violin) were spectrally analyzed and then equalized for duration, attack and decay time, fundamental frequency, number of harmonics, average spectral centroid, and presentation loudness. The tones were resynthesized both with time-varying harmonic amplitudes and frequencies (dynamic case) and fixed amplitudes and frequencies (static case). Tone triads were presented to ten musically experienced listeners whose tasks were to specify the most dissimilar and most similar pairs in each triad. Based on the resulting dissimilarity matrix, multidimensional scaling (MDS) was used to position the instruments in two- and three-dimensional metric spaces. Two measures of instrument amplitude spectra were found to correlate strongly with MDS dimensions. For both the static- and dynamic-case 2-D solutions, the ratio of even-to-odd rms amplitudes correlated strongly with one of the dimensions. For the dynamic case, spectral centroid variation correlated strongly with the second dimension. Also, 2-D solution instrument groupings agreed well with groupings based on coefficients of the first two components of a principle components analysis representing 90% of the instruments’ spectral variance. [This work was supported by the Hong Kong Research Grants Council’s CERG Project 613505.]

Inexperienced Listeners' Perception of Timbre Dissimilarity Within and Between Voice Categories

In this paper, a novel acoustic surveillance system, that can be used in critical facilities and infrastructures, border security and public premises is introduced. As an application, an acoustic surveillance system that detects gunshots from recorded acoustic signals, and identifies the originating weapon type is selected. Mel Frequency Cepstral Coefficients, zero crossing rate, spectral percentile, spectral centroid, spectral spread and spectral flatness are used as features in different classifiers, namely, the Gaussian Mixture Models, k-Nearest Neighbor and Support Vector Machines, and their classification performances are compared. The feature space that is composed of features from four different classes is visually analyzed by using the Principle Component Analysis. For gun type identification, the best classification performance of 80% was achieved using the k-NN classifier.

The two important aspects of speech are quality and intelligibility. Out of the two, intelligibility has direct relation to speech perception. We propose an algorithm for speech enhancement on android platform by Consonant-Vowel-Ratio (CVR) modification. Frequency domain parameters such as spectral centroid and spectral spread are used for the detection of consonant segments in Vowel-Consonant-Vowel (VCV) utterances. Speech enhancement consists of increasing the intensity of consonants relative to the vowel segments by 6dB.

In this paper, an attempt has been made to identify palatal fricative fronting in children speech, where postalveolar /sh/ is mispronounced as dental /s/. In children’s speech, the concentration of energy (darkest part) of spectrogram for /s/ ranges 4000 Hz to 8000 Hz, whereas it ranges 3000 Hz 8000 Hz for /sh/. Gammatonegram follows the frequency subbands of the ear (wider for higher frequencies). Various spectral properties such as spectral centroid, spectral crest factor, spectral decrease, spectral flatness, spectral flux, spectral kurtosis, spectral spread, spectral skewness, spectral slope and Shannon entropy of the spectrogram (interval of 2000 Hz), extracted from the Gammatonegram are proposed for the characterization of /sh/ and /s/. The dataset recorded from 60 native Kannada speaking children of age between 3 1/2 to 6 1/2 years is considered for the analysis from NITK Kids’ Speech Corpus. Support vector machine (SVMs) is considered for the classification. Various combinations of the proposed features are considered for the evaluation, along with the MFCCs(39) and LPCCs(39). Combination of MFCCs(39), LPCCs(39) and Entropy(4) is observed to achieve highest mispronunciation identification performance of 83.2983%.

 Abstract—Cardiovascular diseases (CVDs) are currently the leading cause of deaths worldwide. The traditional auscultation is cost-effective and time-saving for the public to diagnose CVDs early. While many approaches for analysis of the heart sound (HS) signal from auscultation have been utilized successfully, few studies are focused on acoustic perspective to interpret the HS signal. This paper creatively proposes a multidimensional feature extraction technique based on timbre model to interpret HS, which stems from clinical diagnostic basis and medical knowledge. The extracted features have three dimensions, including spectral centroid (SC), log attack time (LT) and temporal centroid (TC). The simulation experiments indicate that the proposed method is promising in HS feature extraction and the later CVD diagnosis.

In this paper we present a new system to classify TV programs into predefined categories based on the analysis of their audio and video contents. This is very useful in intelligent display and storage systems that can select channels and record or skip contents according to the consumer's preference. Distinguishable patterns exist in different categories of TV programs in terms of human faces and audio. In this paper four categories divided into news, cartoon, variety and sport are of interest. News and variety have differences between frames less than sport and cartoon. For audio feature, we apply short time energy, zero crossing, spectral centroid and short time Fourier transform for feature extraction. For face feature, in the first step, Haar like feature is employed for face detection and eigenface is then applied for feature extraction. Then, neural network is implemented for classification. From experimental results, classification rate of 95% accuracy which is better than the other paper is achievable.

In physiological signal analysis, identifying meaningful relationships and inherent patterns in signals can provide valuable information regarding subjects' physiological state and changes. Although MATLAB has been widely used in signal processing and feature analysis, Python has recently dethroned MATLAB with the rise of data science, machine learning and artificial intelligence. Hence, there is a compelling need for a Python package for physiological feature analysis and extraction to achieve compatibility with downstream models often trained in Python. Thus, we present a novel visualization and feature analysis Python toolbox, PySio, to enable rapid, efficient and user-friendly analysis of physiological signals. First, the user should import the signal-of-interest with the corresponding sampling rate. After importing, the user can either analyze the signal as it is, or can choose a specific region for more detailed analysis. PySio enables the user to (i) visualize and analyze the physiological signals (or user-selected segments of the signals) in time domain, (ii) study the signals (or user-selected segments of the signals) in frequency domain through discrete Fourier transform and spectrogram representations, and (iii) investigate and extract the most common time (energy, entropy, zero crossing rate and peaks) and frequency (spectral entropy, rolloff, centroid, spread, peaks and bandpower) domain features, all with one click. Clinical relevance- As the physiological signals originate directly from the underlying physiological events, proper analysis of the signal patterns can provide valuable information in personalized treatment and wearable technology applications.

With the increasing complexity of modern VLSI circuits, achieving high quality built-in self-test requires monitoring an increasingly large number of internal nodes. Due to the limitations in observing large numbers of nodes, it has become increasingly necessary to compact the output from a large number of lines to a small number of lines in a process known as space compaction. In the past, space compaction has usually been performed by circuit-independent compactors, such as the popular parity function. Recently, a class of circuit-specific space compactors, known as programmable space compactors (PSCs), has been proposed. A PSC is a highly effective space compactor tailored for a specific circuit under test (CUT) subjected to a specific set of test stimuli. Circuit-specific information such as the fault-free and expected faulty behaviour of a circuit can be used to choose compactors which have better fault coverage and/or lower area costs than the parity function. The drawback of PSCs is the difficulty involved in finding optimal PSCs for a circuit. The space of possible PSC functions is extremely large and grows exponentially with the number of lines to be compacted. Although searching for optimal PSC functions is a difficult problem, it is not impossible. This thesis presents a method for searching for combinational PSCs based upon genetic algorithms (GAs), a randomized search technique. Randomized search techniques, such as simulated annealing and genetic algorithms, have become increasingly popular in recent years as methods for solving search and optimization problems beyond the capabilities of classical techniques. This thesis develops a procedure for using circuit-specific information in the application of (GAs) to the search for optimal PSCs for a given CUT and a given set of test stimuli. A measure of the effectiveness, or fitness, of PSCs is developed as a means for guiding the genetic search. The factors used to assess the effectiveness of a PSC are its

Assessment of wood quality for Sape making: Vibroacoustic analysis and machine learning classification

For subjective assessment of speech quality in codecs, a reference impairment system is required to introduce controlled degradations to calibrate subjective evaluation. A reference system provides a convenient means for making meaningful comparisons between subjective test results across laboratories and can be viewed as a scale on which mean opinion scores are projected, this scale being supposed to cover the whole range of quality. Nowadays, standardized anchor systems do not fit any more the degradations brought by the present codecs. This paper aims at offering new reference signals simulating the defaults of codecs currently used on telecommunication networks. Twenty wideband codecs are compared through dissimilarity tests. A multidimensional scaling technique allows us to define a four-dimensional perceptive space that appears stable for male and female talkers. A verbalization task suggests qualifying the degradations perceived by the listeners with the following attributes: muffle, background noise, noise on speech, and hiss, each conveyed by one dimension. These dimensions are correlated with objective measures such as spectral centroid, energy in the silent part in the high frequency sub-band, ratio of brightness between deterministic part and residual part of the signal and spectral correlation coefficient. New reference signals are produced and a phase of validation suggests a perceptive space quite coherent with the original one.

Dissimilarity and the Classification of Female Singing Voices: A Preliminary Study