Abstract: In this paper, a supervised approach based on Convolutional Neural Networks (CNN) for polyphonic piano transcription is presented. The system consists of pitch detection model, onset/offset detection model, and note search model. The pitch detection model is a single-channel CNN predicting the probabilities of pitches contained in one frame of the audio. The onset/offset model based on dual-channel CNN is used for estimating the probabilities of each pitch's onset or offset in a frame. The note search model is rule-based; it integrates the outputs of the pitch model and onset/offset model to determine the final onset, offset and pitch of notes in audio. Two experiments with different dataset conditions are accomplished to compare with state-of-the-art approaches on the same datasets. Experimental results reveal that the proposed approach preforms better in both frame- and note-based metrics.

Abstract: Pitch detection is a fundamental problem in speech processing as F0 is used in a large number of applications. Recent papers have proposed deep learning for robust pitch tracking. In this letter, we consider voicing detection as a classification problem and F0 contour estimation as a regression problem. For both tasks, acoustic features from multiple domains and traditional machine learning methods are used. The discrimination power of existing and proposed features is assessed through mutual information. Multiple supervised and unsupervised approaches are compared. A significant relative reduction of voicing errors over the best baseline is obtained—20% with the best clustering method ( K -means) and 45% with a multi-layer perceptron. For F0 contour estimation, the benefits of regression techniques are limited though. We investigate whether those objective gains translate in a parametric synthesis task. Clear perceptual preferences are observed for the proposed approach over two widely used baselines (robust algorithm for pitch tracking (RAPT) and distributed inline-filter operation (DIO)).

Abstract: Speech activity detection (SAD) is an important component for various speech processing applications and has been researched extensively recently. The pitch continuity, a significant characteristic of speech, however, has not successfully played a role in existing SAD methods. In this work, we propose a novel way to integrate the pitch continuity with pitch-related features. Practice is carried out through the Combo-SAD approach: We examine three consecutive frames and assume that they all have the same pitch as the center frame due to pitch continuity. Corresponding feature values are recomputed at the adjusted pitch location and then used in the final expression. The new combo feature is evaluated with various types of additive noise at different signal-to-noise ratios (SNR). The results show that the new feature leads to better SAD performance (with an up to 39.3% relative improvement on miss rate compared to Combo-SAD). We also introduce a novel variant of the underlying autocorrelation function and illustrate how it can improve the accuracy of pitch detection.

Abstract: Autocorrelation functions (ACF) have been used in various pitch detection algorithms (PDA) and voicing-feature based speech activity detection (SAD) techniques. Speech is assumed to be stationary over a short-term window, and a Hanning window is typically applied in the calculation of ACF. As a result of windowing, the ACF tapers as the autocorrelation lags increase. Boersma demonstrated that the tapering effect could be compensated for by dividing the ACF of the windowed signal by the autocorrelation of the windowing function itself, referred to as wACF hereafter. We recently found that wACF could cause overcompensation and therefore, result in errors in pitch detection. In this paper, a novel normalization method, eACF, is proposed that can both mitigate the tapering effect and minimize the overcompensation. The new method is evaluated on synthetic speech and on the TIMIT database with various types of additive noise at different signal-to-noise (SNR) ratios. The results show that the new method leads to better performance both in terms of pitch detection and speech activity detection. In this paper, we also investigate the scenarios where applying the wACF method is advantageous and where it is not.

Abstract: This paper presents an algorithm for fundamental frequency detection on polyphonic acoustic musical signals, based on a new ‘raking’ method over the frequency-domain spectra. The algorithm is evaluated as a classifier, and boasts a good accuracy (83.20%) compared to other such methods, as well as the ability to function effectively in real-time, with a running-speed below 140ms per window evaluated. This proves to be real-time for the use-case, as the latency between an auditory stimulus and its perception by a person has been shown to be longer than this. The algorithm itself runs in linear-time, but is thus slowed by the O(nlog(n)) Fast Fourier Transform during preprocessing. Though the algorithm fails to account for certain edge-cases with overlapping harmonics as well as certain instruments, future work and improvements are also presented, paving the way for further research.

Abstract: Musicians have been repeatedly reported to show remarkable inter-individual differences in elementary hearing functions, sound perception mode, musical instrument preference, performance style, as well as musical abilities such as absolute- and relative pitch perception or auditory imagery (audiation). However, relevant literature in the field regarding perceptual and psychophysical aspects of sound and particularly pitch perception is highly contradictory, and subjective differences are mostly unconsidered. Moreover, it is largely unexplored how individual differences in (musical) pitch perception are related to further musical abilities and behavior. In the present work, “auditory fingerprints” were created basically based on a composite of five psychoacoustic hearing tests assessing subjective pitch perception in musicians. A total of 93 musicians, including 49 professionals and 44 amateurs, were individually measured for: (a) pitch perception preference (holistic vs. spectral mode), (b) relative pitch perception (musical interval recognition), (c) absolute pitch perception („perfect pitch“), (d) frequency discrimination threshold (just noticeable difference), and (e) auditory imagery of tone-sequences. Overall, eight psychoacoustic parameters were extracted and analyzed using statistical methods. In addition, preferences for musical instruments were evaluated. On the individual level, the results show a high inter-individual variability across the eight psychoacoustic parameters, reflecting clear individual differences in pitch perception and related musical abilities. In addition, principal component analysis (PCA) revealed four different main components to sufficiently represent coherent aspects of the psychoacoustic data, namely: tonal musicality, pitch & timbre preference, low-band sensitivity and high-band sensitivity. On the group level, multi-parametric cluster analyses revealed three sub-groups of subjects, showing significantly different results with respect to the underlying perceptional patterns. Consequently, at least three different modes of pitch perception are suggested, characterized by: 1. Pronounced analytic pattern recognition, focused on spectra / timbre and sensitive to single tones; 2. Pronounced holistic pattern recognition, focused on (missing) fundamental pitch and rather insensitive to single tones; 3. Less pronounced audiation and pitch detection abilities, linked to ambiguous multi-pitch sensations (“balanced mode”). Taken together, the findings suggest that individual “auditory fingerprints” extracted from psychoacoustic hearing tests, reflect remarkable inter-individual differences, but also typical patterns of perceiving (musical) pitch and sound. It can be concluded, that specific auditory characteristics are related to the individual musical (instrument) preference, style and performance of musicians, as well their learning abilities.

Abstract: The invention provides a three-coordinate radar plot clotting method based on 2D sliding window local extremum In the range dimension, range detection clotting is carried out based on the M/N criterion In the pitch dimension, pitch detection clotting is carried out through an 'I'-shaped search window with range lattices and pitch beams as units and based on a 2D sliding window local extremum method In the azimuth dimension, azimuth detection clotting is carried out through a 'nine-rectangle-grid' search window based on a 2D sliding window When that the extension of a target in the azimuthwave position ends is determined, final plot clotting information is output According to the method of the invention, through fine detection clotting based on a 2D sliding window in the pitch and azimuth dimensions, the problem that 'missing detection' is caused in azimuth detection clotting as a target cannot be 'aligned' in range and pitch in different azimuth wave positions is solved, and theprobability of target plot detection is improved

Abstract: According to embodiments of the present invention, a musical performance evaluation method utilizes a musical composition database stored on a server containing musical encodings of all notes to be performed in a musical composition. A student user performing requested musical composition for evaluation may use a client application from a client computing device to receive audio input from the performance and compare identified frequencies to intended frequencies, and identify deviations from pitch, rhythm, and tempo throughout the performance. Deviations may be identified by switching between several pitch detection algorithms based upon a type of note (single note, chord, or plucked string) expected to be played by a student user in real time. Factors for deducting from a performance pitch score and a performance rhythm score are determined and used to calculate a performance pitch score and a performance rhythm score. Error data and scores generated may be transmitted to a server and stored in a historical session database. A student user and an instructor user may review the historical session database to determine the student's progress in musical education.

Abstract: Pitch is one of the most essential features in human speech analysis. Although numerous pitch detection methods have been developed, it is still a challenge to provide a high pitch detection performance in noisy environments. In this paper, we propose an anti-noise pitch detection method that combines a speech enhancement algorithm with a spectral flattening algorithm. In the experiments, we compare the proposed method with several widely used or state-of-the-art pitch detection methods. The results show that the proposed method has the lowest gross pitch error (GPE) rate among all the methods when dealing with white-noise added male speeches. Moreover, comparing the pitches estimated by the proposed method to those estimated by the conventional lag window method, we can see that the speech enhancement algorithm helps diminish pitch errors.

Abstract: The utility model discloses a constant temperature digital display pitch extensibility appearance relates to a pitch detection device, aims at solving current constant temperature digital display pitch extensibility appearance and detects the problem that the precision is low, and its technical scheme main points are, sets up the transparent plate that the lot is located basin diapire top at the basin inside seal, is formed with the cavity between transparent plate and the basin diapire, sets up the lamp source in the cavity, be provided with the convex lens that are located the aqueous mediumtop, play amplification effect in the basin, reached and improved the definition of surveing, the effect that improves experimental precision.

Abstract: Real-time pitch detection of voiced musical notes involves converting sound waves, produced by a voiced rendition of one or more musical notes, to a time domain electronic audio signal. The electronic audio signal is processed to determine a true pitch of the time domain electronic audio signal. True pitch information is displayed in real-time, concurrent with the voiced rendition of each musical note. A pitch indicator conveys to a user information concerning the true pitch which has been determined. The true pitch is determined by segmenting the electronic audio signal into a plurality of audio signal samples and applying a constant-Q transform. Additional processing steps are applied to reduce pitch detection errors.

Abstract: Most MFCC based speech recognition algorithms employ frame segmentation to divide a signal into fixed-size frames as the first step prior to MFCC feature extraction. Commonly used fixed frame sizes, around 20-40 ms, do not usually fit into complete periods of speech signals. Consequently, in MFCC feature extraction, spectral leakage arises after Discrete Fourier Transform is applied to these fixed-size intervals resulting in smeared spectra and reduced speech recognition performance. In this paper, a pitch-based speech signal segmentation to reduce spectral leakage is proposed by utilizing a new technique of pitch detection based on Short-time Energy Waveform (SEW) to yield segmented speech intervals with complete periods. The proposed method utilizes local minima of SEW as markers for pitch segmentation. After segmenting speech signals into pitches, MFCC feature vectors are extracted and subsequently used as raw data for speech recognition using artificial neural networks. Speech recognition experiments using artificial neural networks, applied to collect Thai language speech signals from 40 speakers, were conducted. Empirical results indicate that speech recognition using speech signals segmented into pitches yields more accurate recognition results than those using speech signals segmented into a fixed frame.

Abstract: The invention discloses a pitch calibration method and device for a pitch-adjustable propeller and belongs to the technical field of ship propulsion The method comprises the steps of controlling a drive mechanism to drive blades to rotate at a uniform speed; sequentially increasing (decreasing) the pitch corresponding to a pitch adjusting command, obtaining the pitch detected by a pitch detectionsensor and a drive mechanism load detected by a load sensor until the drive mechanism load is greater (smaller) than or equal to the set maximum (minimum) drive mechanism load value every time the pitch corresponding to the pitch adjusting command is increased (decreased), and regarding the pitch detected by the pitch detection sensor when increasing (decreasing) of the pitch corresponding to thepitch adjusting command is stopped as the pitch calibration value of the greatest advancing (reversing) position; and if the drive mechanism load is minimum, regarding the pitch detected by the pitchdetection sensor as the pitch calibration value of the zero-pitch position According to the pitch calibration method and device for the pitch-adjustable propeller, operation is simple and timely

Abstract: a pitch detection unit detecting a pitch on a time-series basis based on the input sound, a sound-volume detection unit detecting a sound volume on the time series basis based on the input sound, a first starting-point detection unit determining whether variation of the sound volume is equal to or larger than a predetermined threshold for each predetermined period and detecting a starting point of a period in which the variation of the sound volume is equal to or larger than the threshold as a first starting point, and a technique determination unit determining a technique of the input sound based on a change of the sound volume after the first starting point and variation of the pitch after the first starting point.

Abstract: Collaboratively creating musical harmonies includes receiving a user selection of a particular harmony. In response to this selection, there is displayed on a display screen of a computing device a plurality of musical note indicators or notes to specify a first harmony part of a musical piece to be performed. Real-time pitch detection is used to determine a pitch of each note which is voiced by a person, and a graphic indication of the actual pitch which is sung is displayed in conjunction with the musical note indicators.

Abstract: This paper proposes a framed-based extended MIDI-pitch detection algorithm based on variable sampling technology. Symbols P--, P-, P+ and P++ are used to present four different logarithmic frequencies around the MIDI pitch P (integer) within 100 cents interval and called as extended MIDI-pitch to enhance the expression resolution of MIDI pitch of a signal frame. Experiment shows that the algorithm has a good raw pitch accuracy rate and less computation complexity o(nlog2n). Keywords—extended MIDI pitch; singing transcription; singing signal processing

Abstract: Fundamental frequency estimation is one of the most important issues in the field of speech processing. An accurate estimate of the fundamental frequency plays a key role in the field of speech and music analysis. So far, various methods have been proposed in the time- and frequency-domain. However, the main challenge is the strong noises in speech signals. In this paper, to improve the accuracy of fundamental frequency estimation, we propose a method for optimal combination of fundamental frequency estimation methods, in noisy signals. In this method, to discriminate voiced frames from unvoiced frames in a better way, the Voiced/Unvoiced (V/U) scores of four pitch detection methods are combined linearly. These methods are: Autocorrelation, Yin, YAAPT and SWIPE. After identifying the Voiced/Unvoiced label of each frame, the fundamental frequency (F 0 ) of the frame is estimated using the SWIPE method. The optimal coefficients for linear combination are determined using the regularized least squares method with Tikhonov regularization. To evaluate the proposed method, 10 speech files (5 female and 5 male voices) are selected from the PTDB-TUG standard database and the results are presented in terms of SDFPE, GPE, VDE, PTE and FFE standard error criteria. The results indicate that our proposed method relatively reduced the aforementioned criteria (averaged in various SNRs) by 27.13%, 22.14%, 17.40%, and 26.74% respectively, which demonstrate the effectiveness of the proposed method in comparison to state-of-the-art methods.

Abstract: Pitch extraction is a well-known and prominent application in signal processing. However, there are chances for improvements yet. This paper proposes a methodology based on laryngeal mechanisms background for pitch extraction. The goal is to optimize the frequency range of the pitch detectors relying on the fact that each laryngeal mechanism has a frequency range, and therefore, it is not necessary to use all range of human voice to every sound, since it was produced in just one specific mechanism. Since our proposal consists of parameter tuning, it is not limited to specific pitch extraction algorithms. Our experiments show that this adjustment on the frequency range can significantly improve pitch detection accuracy.

Abstract: Pitch extraction is one of the most important areas in signal processing. One of the reasons for this fact is because it is a key component of several speech processing, coding or synthesis systems. Many methods were proposed to date; however, there are possible improvements yet, mainly concerning the fine-tuning of parameters, since the majority of the works focus on the definition of completely new approaches. This paper proposes a system to estimate pitch based on laryngeal mechanisms classification. Currently, this classification is based on texture discrimination between visual representations of the audio signal. Therefore, at first, we improve state-of-the-art accuracy in classifying laryngeal mechanisms through changing this image generation process, based on the spectrogram of the signal, and by tuning the used classifier. After that, we optimize the frequency range of the pitch detectors relying on that classification. Such optimization is possible because each laryngeal mechanism has a frequency range, and therefore, it is not necessary to use the complete range of human voice to every sound; it was produced in just one specific mechanism. Since our system consists of parameter tuning, it is not limited to any specific pitch extraction method. Our experiments over two well-known pitch detectors show that this adjustment on the frequency range based on laryngeal mechanism can significantly improve pitch detection accuracy.

Abstract: The pitch detection is an integral element of automatic music transcription system. Empirical Mode Decomposition (EMD) technique plays a key part in the pitch detection. With this technique, any complicated data set comprising of frequency-amplitude points can be decomposed into small number of finite Intrinsic Mode Functions (IMF). The IMF logic is in accordance with a well-behaved and well proven Hilbert transform. In this paper, a step by step algorithm for detecting pitch period from classical music signal based on Hilbert-Huang transform (HHT) is proposed. Traditional windowing methods have two limitations namely overlapping of windows and an assumption of stationary pitch period within a window. In contrast, HHT shows no limitations on window selection and allows pitch period changing within windows. It also can be used to monitor the variation of the pitch. To validate the proposed method, the pure tone of standard pitch is used. The results show that the variation of the pitch period can be accurately detected. This demonstrates the successful application of Hilbert-Huang transform for pitch detection from Indian classical music signal.

Abstract: The invention provides a two-degree-of-freedom motion attitude measurement method based on visual recognition for an airborne radar antenna, and belongs to the field of motion attitude measurement ofairborne radar antennas. The method comprises the following steps: setting an orientation detection identifier on a mechanism for azimuth movement, setting a pitch detection identifier on a mechanismfor azimuth and pitch movement, collecting motion start and end points and a motion time of the detection identifier by an image collection device, calculating a rotation angle according to the distance between the motion start and end points and the motion radius, and calculating angular velocities and angular accelerations for azimuth and pitch according to the rotation angle and the motion time. The method can accurately detect two-degree-of-freedom motion attitude information of the airborne radar antenna within a certain range, and can finally realize the detection of angular velocities and angular accelerations for azimuth and pitch of antennas, which is convenient, accurate and reliable.

Abstract: Most work on automatic transcription produces “piano roll” data with no musical interpretation of the rhythm or pitches We present a polyphonic transcription method that converts a music audio signal into a human-readable musical score, by integrating multi-pitch detection and rhythm quantization methods This integration is made difficult by the fact that the multi-pitch detection produces erroneous notes such as extra notes and introduces timing errors that are added to temporal deviations due to musical expression Thus, we propose a rhythm quantization method that can remove extra notes by extending the metrical hidden Markov model and optimize the model parameters We also improve the note-tracking process of multi-pitch detection by refining the treatment of repeated notes and adjustment of onset times Finally, we propose evaluation measures for transcribed scores Systematic evaluations on commonly used classical piano data show that these treatments improve the performance of transcription, which can be used as benchmarks for further studies

Abstract: Speech coding forms a crucial element in speech communications. An important area concerning it lies in feature extraction which can be used for analyzing Hindustani Classical Music. An important feature in this respect is the fundamental frequency often referred to as the pitch. In this work, the terms pitch and its acoustical sensation, the frequency is used interchangeably. There exists numerous pitch detection algorithms which detect the main/ fundamental frequency in a given musical piece, but we have come up with a unique algorithm for pitch detection using the binning method as described in the paper using appropriate bin size. Previous work on this subject throws light on pitch identification for Hindustani Classical Music. Pitch Class Distribution has been employed in this work. It can be used to identify pitches in Hindustani Classical Music which is based on suitable intonations and swaras. It follows a particular ratio pattern which is a tuning for diatonic scale proposed by Ptolemy and confirmed by Zarlino is explored in this paper. We have also given our estimated of these ratios and compared the error with the above. The error produced by varying the bin size in our algorithm is investigated and an estimate for an appropriate bin size is suggested and tested. The binning algorithm thus helps to segregate the important pitches in a given musical piece.