Using discrete Tchebichef transform on speech recognition

Intra-speaker variations due to perceptually induced stress or emotion adversely affect speech recognition system performance. In this paper, we combine auditory based (Mel frequency cepstral coefficients and linear predictive cepstral coefficients) features and articulatory based (voicedness) features for robust speech recognition. Voicedness features are derived using linear and teager energy operator (TEO) based nonlinear fast Fourier transform (FFT) spectra. Nonlinear properties are analyzed in both the time and frequency domains. In addition, we investigate the sensitivity of all these FFT spectra to stress and observe the performance of individual FFT spectra. The system is tested using stressed speech data from the speech under simulated and actual stress (SUSAS) database. The results show that articulatory based features help to improve the system performance. Furthermore, significant performance improvement has been observed when using the FFT spectrum which is less sensitive to stress

The proliferation of personal artificial intelligence (AI) -assistant technologies with speech-based conversational AI interfaces is driving the exponential growth in the consumer Internet of Things (IoT) market. As these technologies are being applied to keyword spotting (KWS), automatic speech recognition (ASR), natural language processing (NLP), and text-to-speech (TTS) applications, it is of paramount importance that they provide uncompromising performance for context learning in long sequences, which is a key benefit of the attention mechanism, and that they work seamlessly in polyphonic environments. In this work, we present a 25-mm2 system-on-chip (SoC) in 16-nm FinFET technology, codenamed SM6, which executes end-to-end speech-enhancing attention-based ASR and NLP workloads. The SoC includes: 1) FlexASR, a highly reconfigurable NLP inference processor optimized for whole-model acceleration of bidirectional attention-based sequence-to-sequence (seq2seq) deep neural networks (DNNs); 2) a Markov random field source separation engine (MSSE), a probabilistic graphical model accelerator for unsupervised inference via Gibbs sampling, used for sound source separation; 3) a dual-core Arm Cortex A53 CPU cluster, which provides on-demand single Instruction/multiple data (SIMD) fast fourier transform (FFT) processing and performs various application logic (e.g., expectation–maximization (EM) algorithm and 8-bit floating-point (FP8) quantization); and 4) an always-ON M0 subsystem for audio detection and power management. Measurement results demonstrate the efficiency ranges of 2.6–7.8 TFLOPs/W and 4.33–17.6 Gsamples/s/W for FlexASR and MSSE, respectively; MSSE denoising performance allowing 6 <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\times $ </tex-math></inline-formula> smaller ASR model to be stored on-chip with negligible accuracy loss; and 2.24-mJ energy consumption while achieving real-time throughput, end-to-end, and per-frame ASR latencies of 18 ms.

Purpose Speech recognition technology (SRT) is increasingly available and may provide opportunities for individuals with disabilities to participate in necessary activities and meaningful occupations. This inquiry methodically collects and reports on research related to SRT for individuals with disabilities and impairments. Materials and methods Using the Arksey and O’Malley framework and PRISMA guidelines, five databases were queried and indexes of 11 journals were hand-searched for relevant articles. Articles were included if they were published after 2005, involved SRT, written in English, and participants(s) had a disability. Articles were charted, categorised for level of evidence, and findings were summarised. Results and conclusions Of the 78 articles that were retrieved, 13 met inclusion criteria and were organised into 4 categories: SRT in primary and secondary education, in post-secondary education, for daily living, and without a specified context. No included articles considered SRT in vocational contexts. It was determined that SRT is a tool that may improve participation and independence of individuals with disabilities in multiple contexts. Most studies reviewed were non-empirical. Opportunities exist for providers to continue to develop a rigorous body of knowledge for the use of SRT to meet educational, vocational and daily living needs. IMPLICATIONS FOR REHABILITATION Speech recognition technology (SRT) provides a method of access that is widely available in mainstream technology devices such as personal computers, smartphones and mobile devices, and smart speakers/daily living aids. SRT may improve participation and independence of individuals with disabilities in contexts such as primary and secondary education, post-secondary education, and with independent living tasks. Many studies examining SRT are non-empirical, and there is an opportunity for assistive technology professionals and other providers who use assistive technology to continue to develop a rigorous body of knowledge.

In this study, we present initial efforts for a new speech recognition approach aimed at producing different input images for convolutional neural network (CNN)-based speech recognition. We explored the potential of the tympanic membrane (eardrum)-inspired viscoelastic membrane-type diaphragms to deliver audio visualization images using a cross-recurrence plot (CRP). These images were formed by the two phase-shifted vibration responses of viscoelastic diaphragms. We expect this technique to replace the fast Fourier transform (FFT) spectrum currently used for speech recognition. Herein, we report that the new creation method of color images enabled by combining two phase-shifted vibration responses of viscoelastic diaphragms with CRP shows a lower computation burden and a promising potential alternative way to STFT (conventional spectrogram) when the image resolution (pixel size) is below critical resolution.

This paper proposes a Gaussian Mixture Model (GMM)---based speech emotion recognition methods using four feature parameters; 1) Fast Fourier Transform(FFT) spectral entropy, 2) delta FFT spectral entropy, 3) Mel-frequency Filter Bank (MFB) spectral entropy, 4) delta MFB spectral entropy. In addition, we use four emotions in a speech database including anger, sadness, happiness, and neutrality. We perform speech emotion recognition experiments using each pre-defined emotion and gender. The experimental results show that the proposed emotion recognition using FFT spectral-based entropy and MFB spectral-based entropy performs better than existing emotion recognition based on GMM using energy, Zero Crossing Rate (ZCR), Linear Prediction Coefficient (LPC), and pitch parameters.

Classic McEliece has gathered attention as a candidate in NIST post-quantum cryptography standardization. However, it suffers from high demands on the decryption algorithm, making it unsuitable for resource-constrained devices. In this paper, we propose a novel implementation method, named giant footprint sharing, that reduces memory size during decryption in Classic McEliece. The decryption algorithm processes a large number of intermediate variables computed from a secret key in memory. The giant footprint sharing identifies the largest variable among them and allocates a memory-sharing structure to store it, thereby reducing the overall memory size regardless of the implementation platform. The giant footprint sharing can also be combined with existing acceleration techniques, such as fast Fourier transformation. We evaluate Classic McEliece with the giant footprint sharing on the Arm Cortex-M33 CPU and show that it reduces memory size by up to 46% without significant degradation in computation time compared with the existing fast-implementation by Chen et al. (at TCHES 2021). Extensive experiments with the giant footprint sharing further reveal that it maintains a constant memory size regardless of the compiler optimization, and it also achieves an optimal balance in the trade-off between memory size and computation time. The giant footprint sharing is remarkable for any scheme, that contains a large-scale matrix computation and the life cycle for each variable is limited.

With the upcoming fourth generation wireless systems and convergence of multiple radio standards into a single terminal, building blocks are needed that can be configured for computing various algorithms used in different standards. Given these trends and requirements, in our previous paper we successfully proposed a circuit-sharing design for the DAB receiver to integrate the FFT (fast Fourier transform) and IMDCT (inverse modified discrete cosine transform) operations into the same functional circuit. The proposed technique reduces hardware overhead, enhances circuit efficiency and significantly reduces the cost of DAB receivers. To further improve the efficiency of our previous work, this investigation proposes another alternative design named the circuit-sharing pipeline design (CSPD) using a single processor with a pipeline scheme to combine two functions, FFT and IMDCT, into the same circuit. The proposed method reduces the required chip area and cost of DAB receivers. Analyzing the existing relationship among IMDCT, DCT (discrete cosine transform) and FFT, the IMDCT function can be replaced using a FFT function. Therefore, it is not necessary to design an extra circuit for IMDCT. The arithmetic unit in the FFT processor can be significantly reduced due to employing the pipeline scheme. Consequently, the circuit redundancies in the IMDCT and FFT functions can be easily eliminated to allow exploitation of the decreased chip area. Results of this study demonstrate that the proposed design can provide advantages such as low gate count and small memory size in the DAB (digital audio broadcasting) receiver.

Automatic speech recognition (ASR) is very important for mobile devices. However, deep neural network-based ASR demands a large number of computations, while the memory bandwidth and battery capacity of mobile devices are limited. Server-based implementations are mostly employed, but this increases latency or privacy concerns. Efficient on-device ASR is the solution for these issues. In this paper, we propose a low-latency on-device speech recognition system with a simple gated convolutional network (SGCN). The SGCN shows a competitive recognition accuracy even with 1M parameters. In addition, SGCN is advantageous for parallelization which enables efficient cache utilization. 8-bit quantization is applied to reduce the memory size and computation time. The proposed system features online recognition fulfilling the 0.4s latency limit and operates with the real-time factor of 0.2 using only a single 900MHz CPU core. The system occupying 1.2MB memory footprint shows 19.75% word error rate (WER) with greedy decoding.

Accented pronunciation variability is one of the key elements that deteriorate the accuracy of the automatic speech recognition (ASR). This article reports the results of the acoustic analysis of the two groups of speakers’ variability caused by regional accent in Bangladeshi Bangla. The analysis considers the seven monophthongal and four diphthongal vowels of Bangla to investigate the acoustic characteristics of two groups of single-accent speakers and their correlation on the articulation of the Standard Colloquial Bangladeshi Bangla (SCBB). An accent is the speaker’s regional signature and shaped by his/her community and educational background. This study examines both male and female speakers from the Sylhet region, which has one of the extremely deviant dialects in Bangla, and comparatively less deviant speakers from different districts of North-West and Middle Part of Bangladesh. Accent-related acoustic features such as pitch slope, formant frequencies, and vowel duration have been considered to examine the prominent characteristics of the accents and to classify the accents from these features. Both gender groups are distinctly analyzed. It has been found that there are significant deviations in formant frequencies and various steepness of the rise/fall in pitch slope within accents of both gender groups. In this study, it has been observed that accent related changes in speech affect the ASR performance. This has emphasized the need for accent-specific acoustic models to handle the speakers from highly deviant dialects as well as considering the accent-affected speakers’ variability in the corpora development for robust ASR system in Bangladeshi Bangla.

A very efficient approach to using an artificial supervised neural network in Automatic Speech Recognition in the case of speaker dependent continuous utterance is presented in this paper; it has been tested in the Italian language but in principle not limited to it. An automatic segmentation of the digitized signal and a minimum of human intervention was applied to obtain a phoneme recognition efficiency of 98% on Italian phrases constructed with a limited number of 11 alphabetic classes, defining 20 phonetic subclasses. The efficiency we observed is due to the combined effect of four factors: • the differential of the detected signal of the utterance was digitized; • during the parametrization of segments through Fast Fourier Transform (FFT) and critical band definition the effect of a second derivative was simulated: the higher sensitivity in the higher frequency range of ear complex was thus simulated. • the proper input pattern to be used in the early stages of the training of the neural network was selected by a very sensitive similitude algorithm; • a dynamic and repetitive training procedure was applied through which the generalization shown by the network after training was used to modify and select the input patterns as well as to control the number of the output nodes used in successive training.

Fast Fourier transform (FFT) is a key building block for orthogonal frequency division multiplexing (OFDM) systems. Due to the development of wireless portable devices, it is important to minimize the size and power of a FFT processor. One of the methods to satisfy such demands is reducing the size of twiddle coefficient memory. This paper presents an effective coefficient memory reduction scheme for a R2 <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> SDF FFT implementation. When applying a conventional method to an N- point R2 <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> SDF FFT, the number of twiddle coefficients is 3N/4. However, the proposed scheme requires only (N/8+1) coefficients and its additional hardware architecture is very simple. The effectiveness of the proposed method is verified by implementation results on a FPGA.

A portable an distributed expert system has been implemented in order to carry out voice and speech recognition of different people. The device consists of an open source microcontroller connected to a microphone that would be able to instantly recognize words and people, and could be distributed as a real time device in the security field. Two different methods of artificial neural network have been performed to implement the system: the first one has been the Continuous Hopfield Network to store and recover patterns previously stored, in order to know the word used to communicate with the system, the second one has been the Backpropagation Algorithm used to recognize between different speakers. However prior to performing these methods, it has been necessary to pre-process the information so that it can be computed in this type of device, transforming sound files into a set of useful data using a Fast Fourier Transform. The results obtained have been satisfactory with an accuracy above 80% for words recovery with a white noise of 40% of the signal and above 80% for recognizing a person with a new words introduced.

Sounds useful to make a communication among human beings. Where each sound have characteristics and levels of different frequency. Sounds produced by vocal cords with cooperation of all sound-producing organs. Currently, the application of this technology develops quickly and was very useful for human life, for example: helping hearing impaired, making google translate and the others. There are several problems in sounds identifying sound, like velocity and accuracy of processing voice data. While for transform sound signals as input signals into sound frequencies in WAV format required a method of Artificial Neural Network to minimize the problem, using Fast Fourier Transform (FFT) with the help of software Matlab to recognize the power of sound (decibels) and a few other voice characteristics. Design goal is to be able to identifying and analyzing characteristics of sound and voice processing data already saved on the system that have been made so that as you wish. By taking 8 times data on to ten peoples who have recognized sound frequency, the percentage of sound at night more than during day in a row was 75%, 50%, 25%, 37%, and 0% as many as three people, two people, two people, one people, and two people. Differences to this percentage are psychological condition, sound-producing organs and weather conditions on the set of data.

This paper presents an evaluation of bit-width effects in a fixed-point, field-programmable gate array (FPGA)optimized implementation of Mel-frequency cepstral coefficient (MFCC) computation for hidden Markov model (HMM)-based phrase speech recognition, conducted via a software simulation. We developed a software to analyze how the bit widths of fixedpoint operations affect the recognition accuracy of an HMMbased phrase speech recognition system. Experimental results show that recognition accuracy comparable to that of a floatingpoint implementation can be achieved even with significantly reduced bit widths. We determine the minimum required bit widths for key MFCC stages, including the Hanning window and fast Fourier transform (FFT). These findings demonstrate the practicality of the hardware implementation for real-time low-power speech recognition.

The Fast Fourier Transform (FFT) method for solving the 1-D heat diffusion equation offers an efficient approach for resolving partial differential equations (PDEs) with various time steps . FFT is used to transform the 1-D heat diffusion equation into the frequency domain and back to the time domain through inverse FFT. Using mathematical modeling with initial and Dirichlet boundary conditions, the numerical solutions produced by FFT are compared with analytical solutions. The accuracy of the method is validated using MAE and MSE calculated in Matlab. At several time intervals , the obtained MAE and MSE values indicate a good agreement between the numerical and analytical solutions, with very small errors. Numerical stability analysis confirms the reliability of the FFT method across various The variation in time step has a significant impact on the accuracy and stability of the solution. Smaller time steps improve accuracy and stability but require longer computation times. The optimal time step selected in this study is Increasing the number of discretization points also enhances accuracy but implies an increase in computational load and memory usage. The FFT method demonstrates good numerical consistency with increasing