Song Signals Research Articles

Densely Connected Networks with Multiple Features for Classifying Sound Signals with Reverberation.

In indoor environments, reverberation can distort the signalseceived by active noise cancelation devices, posing a challenge to sound classification. Therefore, we combined three speech spectral features based on different frequency scales into a densely connected network (DenseNet) to accomplish sound classification with reverberation effects. We adopted the DenseNet structure to make the model lightweight A dataset was created based on experimental and simulation methods, andhe classification goal was to distinguish between music signals, song signals, and speech signals. Using this framework, effectivexperiments were conducted. It was shown that the classification accuracy of the approach based on DenseNet and fused features reached 95.90%, betterhan the results based on other convolutional neural networks (CNNs). The size of the optimized DenseNet model is only 3.09 MB, which is only 7.76% of the size before optimization. We migrated the model to the Android platform. The modified model can discriminate sound clips faster on Android thanhe network before the modification. This shows that the approach based on DenseNet and fused features can dealith sound classification tasks in different indoor scenes, and the lightweight model can be deployed on embedded devices.

Music Feature Extraction Method Based on Internet of Things Technology and Its Application.

Due to the influence of factors such as strong music specialization, complex music theory knowledge, and various variations, it is difficult to identify music features. We have developed a music characteristic identification system using the Internet-based method. The physical sensing layer of our designed system deploys audio sensors on various coordinates to capture the raw audio signal and performs audio signal processing and analysis using the TMS320VC5402 digital signal processor; the Internet transport layer places audio sensors at various locations to capture the raw audio signal. The TMS320VC5402 digital signal processor is used for audio signal diagnosis and treatment. The network transport layer transmits the finished audio signal to the data base of song signal in the application layer of the system; the song characteristic analysis block in the application layer adopts dynamics. The music characteristic analysis block in the applied layer adopts dynamic time warping algorithm to acquire the maximal resemblance between the test template and the reference template to achieve music signal characteristic identification and identify music tunes and music modes based on the identification results. The application layer music feature analysis block adopts dynamic time regularization algorithm and mel-frequency cepstrum coefficient to achieve music signal feature recognition and identify music tunes and music patterns based on the recognition results. We have verified through experiments, and the results show that the system operates consistently, can obtain high-quality music samples, and can extract good music characteristics.

Extraction of Music Main Melody and Multi-Pitch Estimation Method Based on Support Vector Machine in Big Data Environment.

Main melody extraction and multi-pitch estimation are two important research topics in the MIR field. In this article, the SVM algorithm is used to analyze and discuss music melody extraction and multi-pitch estimation. In the part of multi-fundamental frequency extraction, this article first filters the song signal with equal loudness and weakens the energy of the high-frequency and low-frequency parts of the song signal. Thereafter, the multi-resolution short-time Fourier transform suitable for processing song signals is introduced. In addition, in order to avoid the sharp jump of the estimated melody pitch in the same note duration range, this article proposes a main melody extraction method combining the SVM algorithm with dynamic programming. In this article, more features are used to distinguish the pitch contour of vocal fundamental frequency from that of the nonvocal fundamental frequency, which does not only depend on energy or a certain feature. The experimental results show that the lowest octave error of this method is 1.46. Meanwhile, the recall rate of the algorithm can reach about 95%. This method not only improves the recall rate of the fundamental frequency of the human voice but also improves the recall rate and pitch accuracy rate of the whole main melody extraction system.

Ultrasonic songs and stridulum anatomy of Asiophlugis crystal predatory katydids (Tettigonioidea: Meconematinae: Phlugidini)

ABSTRACT The behavioural ecology of ultrasonic-singing katydids is not well understood, and the general bioacoustics, barely known for a few Neotropical Meconematinae, tends to be overlooked for species from Southeast Asia. These include Asiatic species of Phlugidini, commonly known as crystal predatory katydids. One of its genera, Asiophlugis consists of 16 species for which acoustic signals and stridulum anatomy are broadly unknown. These characters can be used to understand species boundaries. Here, we sampled Asiophlugis from five sites in Malay Peninsula and Borneo Island, recorded the acoustic signals of five species plus one subspecies using ultrasound sensitive equipment, and examined their stridulum anatomy. The calling songs of the taxa involved were documented for the first time. We found that the stridulum anatomy (e.g., tooth distributions, tooth length and tooth density) is distinct between species but less so between subspecies. In contrary, songs of different taxa are different based on acoustic parameters (e.g., pulse duration, peak frequency) and descriptive patterns, even between the subspecies. We also did not observe that song signals are more different in sympatry than in allopatry. Whether this can be generalised requires further sampling, highlighting the need for more research on the ultrasonic acoustic communication in Asiatic katydids.

Singer Recognition Based on Convolutional Deep Belief Networks

Singer recognition is an important branch of music retrieval and classification. This paper focuses on the application of convolutional deep belief networks (CDBN) for singer recognition. First, the system architecture of singer recognition based on CDBN is given, and then the pre-processing of the song signal is described in detail, including sampling, framing, pre-emphasis and windowing. The feature extraction of song signal based on MFCC is described in detail, and the composition and principle of CDBN and its application in singer recognition are introduced. Experiment based on three different feature extraction techniques of LPCC, MFCC and CDBN is carried out and the result is compared and analysed, the experimental results show that CDBN is effective for singer recognition.

A Comparison Study to Identify Birds Species Based on Bird Song Signals

this paper presents a comparison study in automatically identifying bird species based on bird acoustic signals, using audio files from XENO-CANTO online database. The features including Mel-Frequency Cepstral Coefficients (MFCC), geo-related meta-features, and the integration are compared. The learning classifiers Linear Discriminant Analysis (LDA), Support Vector Machine (SVM), k-Nearest Neighbor (kNN), and Ensemble Learning are examined. Our experimental results show that in the comparison study, ensemble learning using discriminant learner with the integration of MFCC features and geo meta-features obtains the best detection performance.

Song learning and cognitive ability are not consistently related in a songbird.

Learned aspects of song have been hypothesized to signal cognitive ability in songbirds. We tested this hypothesis in hand-reared song sparrows (Melospiza melodia) that were tutored with playback of adult songs during the critical period for song learning. The songs developed by the 19 male subjects were compared to the model songs to produce two measures of song learning: the proportion of notes copied from models and the average spectrogram cross-correlation between copied notes and model notes. Song repertoire size, which reflects song complexity, was also measured. At 1year of age, subjects were given a battery of five cognitive tests that measured speed of learning in the context of a novel foraging task, color association, color reversal, detour-reaching, and spatial learning. Bivariate correlations between the three song measures and the five cognitive measures revealed no significant associations. As in other studies of avian cognition, different cognitive measures were for the most part not correlated with each other, and this result remained true when 22 hand-reared female song sparrows were added to the analysis. General linear mixed models controlling for effects of neophobia and nest of origin indicated that all three song measures were associated with better performance on color reversal and spatial learning but were associated with worse performance on novel foraging and detour-reaching. Overall, the results do not support the hypothesis that learned aspects of song signal cognitive ability.

A gain control mechanism governs the weighting of acoustic signal intensity and attractiveness during female decisions

In many acoustical communication systems song signals serve for mate selection. Animals thus need to make a decision about which is the best mating partner in a choice situation by integrating the cues of the available signals. In female crickets, choice behaviour is based on the attractiveness of the temporal pattern of a male's calling song and on its intensity. Signal intensity is affected by both sound amplitude and sound energy as given by the amplitude modulation. Here we targeted the psychometric functions that relate these physical cues to sensory perception in female crickets, Gryllus bimaculatus. First, no-choice tests were conducted to examine the response functions for pattern attractiveness and intensity. Second, we systematically varied pattern attractiveness and sound amplitude or sound energy of two competing signals which were presented from opposite sides in a choice situation. To examine the relative contribution of both cues to female choice behaviour an equivalence function for pattern attractiveness and signal intensity was derived which indicated at which levels both cues compensated for each other. The psychometric data and the equivalence function revealed that for patterns of high attractiveness there was a strong impact of even small intensity differences, whereas for patterns with low attractiveness large intensity differences were required to influence female decisions. Thus the integration of pattern attractiveness and intensity is provided by a gain control mechanism that reflects a nonlinear weighting of intensity differences by pattern attractiveness.

Firing-rate resonances in the peripheral auditory system of the cricket, Gryllus bimaculatus.

In many communication systems, information is encoded in the temporal pattern of signals. For rhythmic signals that carry information in specific frequency bands, a neuronal system may profit from tuning its inherent filtering properties towards a peak sensitivity in the respective frequency range. The cricket Gryllus bimaculatus evaluates acoustic communication signals of both conspecifics and predators. The song signals of conspecifics exhibit a characteristic pulse pattern that contains only a narrow range of modulation frequencies. We examined individual neurons (AN1, AN2, ON1) in the peripheral auditory system of the cricket for tuning towards specific modulation frequencies by assessing their firing-rate resonance. Acoustic stimuli with a swept-frequency envelope allowed an efficient characterization of the cells' modulation transfer functions. Some of the examined cells exhibited tuned band-pass properties. Using simple computational models, we demonstrate how different, cell-intrinsic or network-based mechanisms such as subthreshold resonances, spike-triggered adaptation, as well as an interplay of excitation and inhibition can account for the experimentally observed firing-rate resonances. Therefore, basic neuronal mechanisms that share negative feedback as a common theme may contribute to selectivity in the peripheral auditory pathway of crickets that is designed towards mate recognition and predator avoidance.

Measuring consistency: spectrogram cross-correlation versus targeted acoustic parameters

Several recent studies indicate that the consistency of song or syllable repetitions signals male quality. However, the comparability and efficacy of different measurement methods is not known. Here, I compared two approaches to measuring the consistency of syllable repetitions within the trills of House Wren (Troglodytes aedon) songs. In the first approach, I calculated the coefficient of variation (CV) in standard time–frequency measurements within each trill. In the second approach, I used spectrogram cross-correlation (SPCC), which measures the maximum pixel-by-pixel similarity of two spectrograms. The two approaches gave correlated estimates of trill consistency, but SPCC was more strongly related to two putatively biologically relevant traits: specifically, SPCC differed more strongly between age classes, and more variation in SPCC could be attributed to individual differences. CV-based measures complemented SPCC measures by clarifying some of the specific acoustic features whose consistency changed with age. Although additional comparisons between measurement approaches would be useful to assess generality, it appears that researchers interested in song or trill consistency should consider using an SPCC, or a combined SPCC- and CV-based, approach.

An Insect Song Signal Feature Extraction Method Based on the Wavelet Packet Analysis

The wavelet packet decomposition method was used to two common insects song signal. Frequency decomposition and feature extraction were made, the feature vectors, eigenvalues and the sound quality evaluation parameter vectors were constructed, then the correlation analysis calculation were made between the eigenvalues and the sound quality evaluation parameter vectors. The results show that the correlation coefficients are good, the average correlation coefficients of cricket and grasshopper song signals are 0.8875 and 0.6942, the results of cricket is much better than grasshopper, it proved that the proposed algorithm is more suitable for cricket song signals analysis, a new and effective sound quality evaluation method for typical insect with friction sound mechanism is provided.

The Correlation Research of Insects Song Sound Quality and Nano-Mechanical Characteristics

The paper proposed a nanomechanical measurements method according to insect parts biometric friction sound, the nanomechanical measurements were carried out on the cricket and grasshoppers wings, meanwhile, two common insects song signal were collected and the sound quality evaluation parameter vectors were constructed from it. The correlation analysis calculations were made between the nanomechanical properties parameters and the sound quality evaluation parameter vectors. The results show that the nanomechanical properties of biological materials have relevance with the sound quality characteristics, and the average correlation coefficient of the cricket is better, the crickets elastic modulus correlation coefficient with loudness were with the highest average correlation coefficient.

A Systematic Approach to Authenticate Song Signal Without Distortion of Granularity of Audible Information (ASSDGAI)

The major concern to authenticate an audio signal is restoring its basic audible properties and providing such a way that may not produce any unusual perception to listeners. Therefore, authenticating an audio signal needs to embed as less as minimum security code within tolerable range of perception. In this paper, perceptual encoding has been applied for structuring song signal in a specified manner with less modification of its amplitude values and followed by embedding formatted data in skewed pattern that will use for verifying any type of alternation by comparing its own structural format and systematic structure of song signal with neighbouring sampled values. A comparative study has been made with similar existing techniques and experimental results are also supported with mathematical formula based on Microsoft WAVE (.wav) stereo sound file.

Voice, Body and the Transmission of the Real in Documentary Theatre

This article analyses how listening is used to develop performances in Alecky Blythe's verbatim theatre. Listening includes Blythe's use of recorded oral interviews for devising performances, and also the actors’ creation of performance by precisely imitating an interviewee's voice. The article focuses on listening, speaking and embodiment in London Road, Blythe's recent musical play at London's National Theatre, which adopted and modified theatre strategies used in her other plays, especially The Girlfriend Experience and Do We look Like Refugees. The article draws on interviews with performers and with Blythe herself, in its critical analysis of how voice legitimates claims to authenticity in performance. The work on Blythe is contextualised by brief comparative analyses. One is Clio Barnard's film The Arbor, a ‘quasi-documentary’ on the playwright, Andrea Dunbar which makes use of an oral script to which the actors lip-sync. The other comparator is the Wooster Group's Poor Theater, which attempts to recreate Grotowski's Akropolis via vocal impersonation. The article argues that voice in London Road both claims and defers authenticity and authority, inasmuch as voice signifies presence and embodied identity but the reworking of speech into song signals the absence of the real. The translation of voice into written surtitles works similarly in Do We Look Like Refugees. Blythe's theatre, Barnard's film and The Wooster Group's performances are a useful framework for addressing questions of voice and identity, and authenticity and replication in documentary theatre. The article concludes by placing Blythe's oral texts amid current debates around theatre's textual practices.

Preservation of IPR of Songs through Embedding Secret Song (IPRSESS)

Abstract Security is one of the major challenging issues for creative industries all over the world. This paper proposes a technique to provide security to digital song by embedding another strip of song as secret code into original song without affecting its a udible quality. Separating phases and amplitudes of both song signals is precursor to the proposed technique followed by shifting of phases of secret song with modulating amplitude to fabricate the authenticating code. The embedded unique secret audio is use d to detect and identify the originality of song from similar available songs. A comparative study has been made with existing techniques and experimental results are also presented based on Microsoft WAVE (“.wav”) stereo sound file.

Enhancing Security of Quality Song with Embedding Encrypted Hidden Code in Tolerance Level (SQHTL)

Embedding secret information with song signal may hamper its audible quality as well as originality also would be infringed. Therefore, passing secret code with song signal precisely needs to measure the embedded data volume with song signal within embedding tolerance level. In this paper, embedded secret data with linear coding principles has been applied, enhanced security criteria has been maintained with the help of elliptic curve cryptographic application. Hiding information into audio signals and audible quality is maintained only by calculating the acceptance ratio of embedding data using human perception, modulation of channel capacity in modified form is used to determine the embedded data and song signal trade-off ratio for getting future guideline embedding information in song signal with correlation among embedded data is fabricated. A comparative study has been made with similar existing techniques for performance analysis and experimental results are also supported with mathematical formula based on Microsoft WAVE (.wav) stereo sound file.

The Steppengrille (Gryllus spec./assimilis): Selective Filters and Signal Mismatch on Two Time Scales

In Europe, several species of crickets are available commercially as pet food. Here we investigated the calling song and phonotactic selectivity for sound patterns on the short and long time scales for one such a cricket, Gryllus spec., available as “Gryllus assimilis”, the Steppengrille, originally from Ecuador. The calling song consisted of short chirps (2–3 pulses, carrier frequency: 5.0 kHz) emitted with a pulse period of 30.2 ms and chirp rate of 0.43 per second. Females exhibited high selectivity on both time scales. The preference for pulse period peaked at 33 ms which was higher then the pulse period produced by males. Two consecutive pulses per chirp at the correct pulse period were already sufficient for positive phonotaxis. The preference for the chirp pattern was limited by selectivity for small chirp duty cycles and for chirp periods between 200 ms and 500 ms. The long chirp period of the songs of males was unattractive to females. On both time scales a mismatch between the song signal of the males and the preference of females was observed. The variability of song parameters as quantified by the coefficient of variation was below 50% for all temporal measures. Hence, there was not a strong indication for directional selection on song parameters by females which could account for the observed mismatch. The divergence of the chirp period and female preference may originate from a founder effect, when the Steppengrille was cultured. Alternatively the mismatch was a result of selection pressures exerted by commercial breeders on low singing activity, to satisfy customers with softly singing crickets. In the latter case the prominent divergence between male song and female preference was the result of domestication and may serve as an example of rapid evolution of song traits in acoustic communication systems.

Calling song signals and temporal preference functions in the cricket Teleogryllus leo

The acoustic display of many cricket species consists of trains of pulses (chirps) with intermittent pauses. Here, we investigated the temporal cues that females of the cricket Teleogryllus leo used to detect a pulse and a chirp pattern on two different time scales. For both patterns, females accepted a wide range of combinations that covered the respective pulse and chirp parameters in the songs of males. In tests with a continuous series of pulses at different modulation frequencies, the transfer function of pattern discrimination was also determined. Females exhibited two ranges of high response scores indicating two temporal filters with an inhibitory interaction. For the modulation frequency of the pulse pattern, the peak of the preference function was rather sharply tuned and at a lower pulse rate than produced by males. These results show that the combined output of both filters did not increase selectivity, but rather enlarged the accepted range of signals.

Generating Audio Signal Authentication through Secret Embedded Self Harmonic Component (GASASH)

In this paper, an algorithm has been proposed to provide authentication of audio signal through concealment of a selective segment of its harmonic components and hiding the selective harmonic components in such a manner that will not affect over its audible quality. The replica of inner harmonic components will use for hidden signature in wavelet domain. Sampling the hidden harmonic components with the help of amplitude coding for generating lower magnitude values is the first phase of proposed technique followed by fabrication of authenticating code by embedding the lower magnitude values with selected coefficients of song signal generated via wavelet transform [symmetrization mode]. The embedded hidden self component of song signal is used to detect and identify the original song from similar available songs with the help of comparing its hidden code and respective similar inner part. A comparative study has been made with similar existing techniques and experimental results are also supported with mathematical formula based on Microsoft WAVE (.wav) stereo sound file.

Enhancing Security of Quality Song With Embedding Encrypted Hidden Code in Tolerance Level (SQHTL)

Embedding secret information with song signal may hamper its audible quality as well as originality also would be infringed. Therefore, passing secret code with song signal precisely needs to measure the embedded data volume with song signal within embedding tolerance level. In this paper, embedded secret data with linear coding principles has been applied, enhanced security criteria has been maintained with the help of elliptic curve cryptographic application. Hiding information into audio signals and audible quality is maintained only by calculating the acceptance ratio of embedding data using human perception, modulation of channel capacity in modified form is used to determine the embedded data and song signal trade-off ratio for getting future guideline embedding information in song signal with correlation among embedded data is fabricated. A comparative study has been made with similar existing techniques for performance analysis and experimental results are also supported with mathematical formula based on Microsoft WAVE (.wav) stereo sound file.