Exploring Hybrid Techniques for Enhanced Pitch Estimation in Speech Processing

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.

Many speech-related works employ the pitch period as a crucial component. Speech signals are typically collected in challenging noisy settings for real-world projects. Therefore, it is now more important than ever for the algorithm to be noise resistant in order to estimate pitch accurately. However, when dealing with noisy speech files at a low signal-to-noise ratio (SNR) value, many state-of-the-art algorithms are unable to produce satisfactory results. In this work, a new noise-resistant pitch estimation algorithm based on spectral subtraction is presented, which uses a weighted function to lessen the impact of the vocal tract effect. Furthermore, to enhance the correlation between the pitch estimates and smoothen the pitch contours, we employ a weighted function that combines the spectral subtraction-based technique as the numerator and the circular average magnitude difference function (CAMDF) as the denominator. We have utilized two noisy speech databases using seven different kinds of recorded ambient noise, and we evaluated our system against three cutting-edge methods. The suggested method lowers the Gross Pitch Error (GPE) rate at practically all SNRs in white noise and performs best on the NTT and KEELE databases.

This paper describes a method for noise-proof detection of the fundamental frequency of the voice in a noisy environment. Noise reduction techniques have been required in the development of a hearing aid, because noise makes intelligibility of hearing awfully inferior. In various methods of noise reduction, the fundamental frequency is often a significant parameter, but it is difficult to extract the frequency from the noisy voice. In order to utilize a comb filter method for noise reduction, a new method of detecting the fundamental frequency is developed by using the property of continuity in the fundamental frequency and the power spectrum envelope (PSE) of the human voice. The continuity of the PSE is utilized for determining the most reliable frequency. The gross pitch error (GPE) is reduced by the determination. Besides the frequency used for the comb filter is obtained from a linear predicting frequency and the latest fundamental frequency from the noisy voice, so as to suppress fluctuation of the frequency that degrades filtered voice. The procedure improves a fine pitch error (FPE) within 5%. The results of the evaluation showed that the present method proved to be superior to the traditional cepstrum method in the GPE and the FPE. We conclude that the proposed frequency detection method is available for noise reduction in the comb filter method.

Robust pitch estimation is important in many areas of speech processing. In voice pathology, diverse statistics extracted form pitch estimation were commonly used to test voice quality. In this study, we compared several established pitch detection algorithms (PDAs) for verification of adequacy of the PDAs. In the database of total pathological voices of 99 and normal voices of 30, an analysis of errors related with pitch detection was evaluated between pathological and normal voices, or among the types of pathological voices. Pitch errors of all PDAs used in this study more or less showed some changes between pathological and normal voices. According to the results of pitch errors, gross pitch error showed some increases in cases of pathological voices; especially excessive increase in PDA based on nonlinear time-series. In an analysis of types of pathological voices classified by aperiodicity and the degree of chaos, the more voice has aperiodic and chaotic, the more growth of pitch errors increased. Consequently, it is required to survey the severity of tested voice in order to obtain accurate pitch estimates.

The field of automatic speech recognition (ASR) has enjoyed more than 30 years of technology advances due to the extensive utilization of the hidden Markov model (HMM) framework and a concentrated effort by the speech community to make available a vast amount of speech and language resources, known today as the Big Data Paradigm. State-of-the-art ASR systems achieve a high recognition accuracy for well-formed utterances of a variety of languages by decoding speech into the most likely sequence of words among all possible sentences represented by a finite-state network (FSN) approximation of all the knowledge sources required by the ASR task. However, the ASR problem is still far from being solved because not all information available in the speech knowledge hierarchy can be directly integrated into the FSN to improve the ASR performance and enhance system robustness. It is believed that some of the current issues of integrating various knowledge sources in top-down integrated search can be partially addressed by processing techniques that take advantage of the full set of acoustic and language information in speech. It has long been postulated that human speech recognition (HSR) determines the linguistic identity of a sound based on detected evidence that exists at various levels of the speech knowledge hierarchy, ranging from acoustic phonetics to syntax and semantics. This calls for a bottom-up attribute detection and knowledge integration framework that links speech processing with information extraction, by spotting speech cues with a bank of attribute detectors, weighting and combining acoustic evidence to form cognitive hypotheses, and verifying these theories until a consistent recognition decision can be reached. The recently proposed automatic speech attribute transcription (ASAT) framework is an attempt to mimic some HSR capabilities with asynchronous speech event detection followed by bottom-up knowledge integration and verification. In the last few years, ASAT has demonstrated good potential and has been applied to a variety of existing applications in speech processing and information extraction.

Speech is the audible acoustic signal generated by the articulatory system (lungs, vocal folds, vocal tract, tongue, lips) to communicate language. The fundamental frequency (F_0) is the lowest frequency component of a speech waveform and corresponds to the vibration rate of the vocal folds during voiced speech. It also determines the pitch of the speaker’s voice. In speech signal processing, this acoustic waveform is captured and analyzed to extract information about what is being said, how it is being said and who is saying it. In various speech processing applications such as voice synthesis, speaker recognition and emotion analysis accurate extraction of the fundamental frequency (F_0) is a vital task. However, the vocal tract’s formants can sometimes significantly alter the glottal waveform’s shape, making it difficult to identify the true pitch. Additionally, in the presence of background noise, traditional pitch detection techniques often experience a considerable decline in performance. This work proposes a robust method for extracting the fundamental frequency by utilizing the complementary advantages of the power spectrum and logarithmic spectrum in noisy speech environments. The power spectrum mitigates noise effects, while the logarithmic operation effectively separates vocal tract characteristics from the source excitation. The proposed approach integrates autocorrelation-based power spectral analysis with cepstral techniques derived from the log spectrum to improve pitch estimation accuracy under adverse conditions. Experimental results on noisy speech datasets show that the proposed hybrid method achieves lower gross pitch error and greater robustness than traditional methods such as BaNa, autocorrelation and cepstral techniques.

The fundamental frequency (F 0) represents pitch in speech that determines prosodic characteristics of speech and is needed in various tasks for speech analysis and synthesis.Despite decades of research on this topic, F 0 estimation at low signal-to-noise ratios (SNRs) in unexpected noise conditions remains difficult.This work proposes a new approach to noise robust F 0 estimation using a recurrent neural network (RNN) trained in a supervised manner.Recent studies employ deep neural networks (DNNs) for F 0 tracking as a frame-by-frame classification task into quantised frequency states but we propose waveform-tosinusoid regression instead to achieve both noise robustness and accurate estimation with increased frequency resolution.Experimental results with PTDB-TUG corpus contaminated by additive noise (NOISEX-92) demonstrate that the proposed method improves gross pitch error (GPE) rate and fine pitch error (FPE) by more than 35 % at SNRs between -10 dB and +10 dB compared with well-known noise robust F 0 tracker, PEFAC.Furthermore, the proposed method also outperforms state-of-the-art DNN-based approaches by more than 15 % in terms of both FPE and GPE rate over the preceding SNR range.

Purpose of The Study. To investigate the effects of an 8-week plyometric depth jump training program on badminton smash accuracy in amateur athletes and examine the relationship between lower body power development and overhead striking precision. Material and methods. Twenty-four amateur badminton athletes (14 males, 10 females; age: 19-25 years) were randomly assigned to experimental (n=12) and control (n=12) groups. The experimental group performed progressive plyometric depth jump training three times per week in addition to regular badminton practice, while the control group maintained only regular practice. Platform heights progressed from 40-55 cm over 8 weeks. Smash accuracy, vertical jump performance, ground contact time, and movement efficiency were assessed pre- and post-intervention. Results. The experimental group demonstrated significant improvements in smash accuracy (29.6%, p < 0.001, d = 2.21), vertical jump height (13.2%, p < 0.001, d = 1.78), and ground contact time (15.7% reduction, p < 0.001). Strong correlations were observed between vertical jump improvements and accuracy enhancements (r = 0.76, p < 0.001). The control group showed minimal changes (4.4% improvement in accuracy, p = 0.089). No significant gender differences were found in adaptation rates (p = 0.085). Conclusions. An 8-week plyometric depth jump training program significantly improves badminton smash accuracy in amateur athletes, suggesting that enhanced lower body power development directly contributes to improved striking precision. The findings support integrating structured plyometric training into regular badminton practice for technical skill enhancement.

Pitch estimation plays an important role on speech processing such as speech coding, synthesis, recognition, and so on. Although current pitch estimation method performs well under clean condition, the performance deteriorates significantly in noisy environment. For this reason, robust pitch estimation against additive noise is required. Modified auto-correlation method is commonly used as pitch estimation in which LP residual is used to compute the auto-correlation. We have previously proposed pitch estimation methods based on Time-Varying Complex AR (TV-CAR) analysis whose criterion is the weighted correlation of the complex residual obtained by the TV-CAR analysis, sum of the harmonics for the complex residual spectrum, or so on. On the other hand, Azarov et al. have proposed an improved method of RAPT (Robust Algorithm for Pitch Tracking) using an instantaneous harmonics that is called IRAPT (Instantaneous RAPT). The IRAPT can perform better estimation than RAPT. Since IRAPT uses band-limited analytic signal to obtain harmonic frequencies, the complex residual signal obtained by the TV-CAR analysis can also be applied to the IRAPT. In this paper, novel pitch estimation method using the instantaneous frequency based on the robust WLP (Weighted Linear Prediction) TV-CAR residual is proposed and evaluated.

Combined speech enhancement and auditory modelling for robust distributed speech recognition

Pitch estimation is still an open issue in contemporary signal processing research. Nowadays, growing momentum of machine learning techniques application in the data-driven society allows for tackling this problem from a new perspective. This work leverages such an opportunity to propose a refined Instantaneous Frequency and power based pitch Estimator method called IFE. It incorporates deep neural network based pitch estimation with audio front end used for extraction of instantaneous frequency and power of signal components. A thorough results analysis is performed and major advantages and shortcomings of this method are identified, leading to a wide array of suggestions for future improvement While IFE exhibits an instantaneous temporal resolution, a comparison is made against state-of-the-art pitch estimators operating on time windows, proving a comparable degree of prediction accuracy (up to 6% accuracy improvement) while maintaining the advantage of higher temporal resolution.

This paper discusses and experiments on frame-by-frame speech signal processing and recognition for Field Programmable Gate Array (FPGA) devices. The system proposes applications including a voice conversion system that needs signal processing and speech recognition for each frame because it requires real-time processing at each frame. Owing to the processing speed, the authors propose algorithms for FPGA as a hardware processor for Voice Activity Detection (VAD) and speech recognition decoder. However, resources for FPGA devices as gate circuits are minimal, therefore, the algorithms need to be customized in order to implement the FPGA. The algorithms are customized for VAD using a 2nd-order autocorrelation function, and for speech recognition using Euclidian distance. These methods implement an FPGA emulator that demonstrates VAD of speech and noise sections and a speech recognition experiment for discriminating Japanese vowels.

An important drawback affecting most of the speech processing systems is the environmental noise and its harmful effect on the system performance. Examples of such systems are the new wireless communications voice services or digital hearing aid devices. In speech recognition, there are still technical barriers inhibiting such systems from meeting the demands of modern applications. Numerous noise reduction techniques have been developed to palliate the effect of the noise on the system performance and often require an estimate of the noise statistics obtained by means of a precise voice activity detector (VAD). Speech/non-speech detection is an unsolved problem in speech processing and affects numerous applications including robust speech recognition (Karray and Marting, 2003; Ramirez et al. 2003), discontinuous transmission (ITU, 1996; ETSI, 1999), real-time speech transmission on the Internet (Sangwan et al., 2002) or combined noise reduction and echo cancellation schemes in the context of telephony (Basbug et al., 2004; Gustafsson et al., 2002). The speech/non-speech classification task is not as trivial as it appears, and most of the VAD algorithms fail when the level of background noise increases. During the last decade, numerous researchers have developed different strategies for detecting speech on a noisy signal (Sohn et al., 1999; Cho and Kondoz, 2001; Gazor and Zhang, 2003, Armani et al., 2003) and have evaluated the influence of the VAD effectiveness on the performance of speech processing systems (Bouquin-Jeannes and Faucon, 1995). Most of the approaches have focussed on the development of robust algorithms with special attention being paid to the derivation and study of noise robust features and decision rules (Woo et al., 2000; Li et al., 2002; Marzinzik and Kollmeier, 2002). The different VAD methods include those based on energy thresholds (Woo et al., 2000), pitch detection (Chengalvarayan, 1999), spectrum analysis (Marzinzik and Kollmeier, 2002), zero-crossing rate (ITU, 1996), periodicity measure (Tucker, 1992), higher order statistics in the LPC residual domain (Nemer et al., 2001) or combinations of different features (ITU, 1993; ETSI, 1999; Tanyer and Ozer, 2000). This chapter shows a comprehensive approximation to the main challenges in voice activity detection, the different solutions that have been reported in a complete review of the state of the art and the evaluation frameworks that are normally used. The application of VADs for speech coding, speech enhancement and robust speech recognition systems is shown and discussed. Three different VAD methods are described and compared to standardized and

Accurate robust pitch tracking in multispeaker environments is an important issue in monaural speech separation. This research presents a spectrotemporal multipitch algorithm designed to detect voiced unvoiced regions in a mixture of multiple speakers, to identify the number of speakers in voiced regions, and to yield the pitch estimates of voiced speakers. Pitch estimation is based on identifying the minima of the multidimensional (MD) AMDF. The number of voiced speakers is determined by the temporal evolution of the MD AMDF. The algorithm has been developed and tested for the two‐speaker case and is being extended to the three‐speaker case. Evaluation on a framewise basis for the two‐speaker case yielded an insertion error rate of 1.78%, a deletion error rate of 14.37%, and a substitution error rate of 3.26%, comparable to the state of the art. Pitch estimation errors resulted primarily from speaker domination. The extension to the three‐speaker case shows promising preliminary results. The proposed algorithm can contribute to speaker separation by helping to identify spectrotemporal regions dominated by one of the speakers from those that have comparable contribution from all speakers. Appropriate strategies can then be used to separate speech in these different scenarios. [Research supported by NSF Grant BCS‐0519256].

This paper introduces a speech separation system as a front-end processing step for automatic speech recognition (ASR). It employs computational auditory scene analysis (CASA) to separate the target speech from the interference speech. Specifically, the mixed speech is preprocessed based on auditory peripheral model. Then a pitch tracking is conducted and the dominant pitch is used as a main cue to find the target speech. Next, the time frequency (TF) units are merged into many segments. These segments are then combined into streams via CASA initial grouping. A regrouping strategy is employed to refine these streams via amplitude modulate (AM) cues, which are finally organized by the speaker recognition techniques into corresponding speakers. Finally, the output streams are reconstructed to compensate the missing data in the abovementioned processing steps by a cluster based feature reconstruction. Experimental results of ASR show that at low TMR (<-6dB) the proposed method offers significantly higher recognition accuracy. Index term: speech separation, CASA, speaker recognition, pitch tracking, units grouping, reconstruction