Questions tagged [speech-processing]

Speech processing is the study of speech signals and the processing methods of these signals.

74 questions with no upvoted or accepted answers
Filter by
Sorted by
Tagged with
4
votes
2answers
428 views

Matching LPC frequency response to raw sample data frequency response

Background: The intention is feature/formant extraction from speech input. I have used an FFT on raw samples in the past and have done simple peak picking from a weighted average of frequencies over ...
3
votes
0answers
62 views

Frame lookup using audio

I'm an animator trying to get a program that will use machine learning to automate the lip syncing process. The program I need made will have two files given to it: a seed, and an audio file. The seed ...
3
votes
1answer
505 views

How to reduce synthesis artifacts produced by phase vocoder?

I implemented the phase vocoder algorithm in Python to time-stretch speech signals by following this paper and referring to this MATLAB tutorial. I can distinguish words in the original signal from ...
3
votes
1answer
295 views

How to prevent Octave Jumping in ACF of Speech Signal?

I'm working with speech signals and my aim is to estimate the fundamental frequency $\ F_0$ of this signal often called as "pitch". The main idea is taking small blocks of the speech signal such ...
3
votes
0answers
1k views

Pre-emphasizing in speech recognition

Usually, in speech recognition, the techniques that are used are based on the linear prediction model (Fant, 1960). Parallel to this, the human speech production mechanism causes energy to drop ...
3
votes
0answers
12k views

MFCC calculation

I've got a project and need to calculate MFCCs. I tried to read some tutorials and then make a MATLAB function but I seem to have wrong answers. I'd like to feed MFCCs to one of the classification ...
2
votes
1answer
364 views

What are F0 counters in Speech signals? What is represented by F0?

I can't find any dedicated definition for Frequency F0 counters. Can anyone tell me what does F0 exactly mean, I am confused about that.
2
votes
1answer
542 views

MFCC Classification

I'm working on gender estimation from speech signal and I completed MFCC feature extraction. So now I'm trying to estimate gender from these features. But I have frames for an audio file and I ...
2
votes
0answers
126 views

Speaker diarization tool for acoustic analysis

I'm researching certain acoustic prosodic features during live clinical interactions. Because of the setting, the recordings always have cross channel bleed, and in some cases everything is recorded ...
2
votes
0answers
140 views

Cepstrum of finite impulse train

In Acero's Spoken language processing (p. 309) it says that the cepstrum of a finite length impulse train $$x[n] = \sum^{M-1}_{k=0} \delta[n-kN]$$ is given by $$\hat{x}[n] = \sum^{ \infty}_{r=1} \frac{...
2
votes
0answers
197 views

Speech features that can be inverted?

I am trying to figure out speech features that can be inverted to recover the audio back with no loss or as less loss as possible. I am currently using the power spectrum which is completely ...
1
vote
0answers
34 views

Logarithmic Amplitude Spectral Subtraction

I'm diving into audio processing and I'm trying to wrap my head around spectral subtraction. I learned there are different approaches do it based on power, magnitude, oversubtraction. My task is to ...
1
vote
1answer
27 views

How to compare 2 MFCC (compare them if they where created from same speaker)

There are multiple speakers. Each speaker generates multiple MFCC. If I'm getting two different MFCC, can we know if those MFCC came from 2 different speakers or same speaker ? How can we compare ...
1
vote
1answer
28 views

Formant analysis LPC (Matlab)

I wanted to do formant analysis using MATLAB and did everything according to the following tutorial: https://de.mathworks.com/help/signal/ug/formant-estimation-with-lpc-coefficients.html I have speech ...
1
vote
0answers
43 views

Understanding MFCC

I am not from electrical or electronic background so my knowledge will be lacking. MFCC is represented by 39 values for each window frame. 12 values are the mel filter-bank and we get 13th value by ...
1
vote
0answers
34 views

Is fundamental frequency information present in speech-shaped white noise?

I am using a so-called speech-shaped stationary white noise, which means so much as a white noise that has been filtered to match the long-term spectrum of a female speaker. The way it was generated ...
1
vote
0answers
727 views

How to overcome different MFCC coefficients distributions from different speech datasets?

I am using MFCC values as features for a machine learning model of speech, detecting age from a voice recording of a person. I work with voice datasets I found on the web: common-voice and vctk. ...
1
vote
0answers
1k views

Python scipy.signal.wiener filter for speech processing

As far as I know Wiener filter can be used in Image processing.I wonder that if i can use Python scipy.signal.wiener() for speech processing as well. Thanks in advance.
1
vote
2answers
1k views

Quantify the difference between two audio spectrograms

I have two speech signals of the same length. They contain the exact same speech content but are recorded from two different microphones. I want to compare the differences between them. 1) Assuming ...
1
vote
1answer
459 views

autoregressive moving average code implementation

I am new to DSP and i am trying to take a wav (human speech) file and apply ARMA filter and plot its PSD graph in python. I see that there are a lot of AR implementations but almost none ARMA. I ...
1
vote
0answers
218 views

What is advantage of Logarithmic Sampling?

Logarithmic Fourier Transformation This paper starts off in the abstract saying that the given audio signal is sampled logarithmically. We also know that the human ear has a sort of a filter bank ...
1
vote
0answers
189 views

What is logarithmic mapping in a filter bank such as the one in the human ear?

"The spacing of the filters in a filter bank used to make an ordinary spectrogram is linear, whereas an auditory filter bank has a spacing that corresponds to the way in which sinusoidal frequency ...
1
vote
0answers
217 views

difference between logged power spectrum and power spectrum?

What is the difference between a logged power spectrum and power spectrum... Well the log part. So can a normal power spectrum be converted to a logged power spectrum just by taking the log of the ...
1
vote
0answers
57 views

Finding the closest match of a recorded word to a 'corpus' of singly-recorded words.

I am interested in learning more about processing audio and thought that it will be helpful to learn while doing a project. The project involves a pseudo-speec recognition where I have a corpus of ...
1
vote
0answers
152 views

Features for speaker recognition

I'm trying to create my own program for speaker identification (text-independent) and have used MFCC features for that. But result of recognition is very poor. My supervisor told me that using MFCC ...
1
vote
0answers
461 views

audio features extraction using restricted boltzmann machine

I want to extract Audio Features using RBM (Restricted Boltzmann Machine). For this, I am giving the spectrogram (PCA whitened) as an input to the RBM. For each audio file, The spectrogram is a matrix ...
1
vote
0answers
252 views

Reducing frequency of a signal

I have a signal that I want to make it sound like a whisper. This is probably better than using a voice conversion system, but I was just want to do some experimenting before. I was thinking maybe ...
1
vote
0answers
36 views

Examples of cases where “invariance to transformation” on signal is important

I realize that this might not be the most well-formulated question, but I figured I'd throw it out there as an open-ended discussion that could turn up some interesting input. I'm relatively new to ...
0
votes
0answers
16 views

What is the difference between MFCC features and Mel Filter Banks (FBANKs)?

I haven't been able to find any source explaining what the FBANK features are. As a reference, I have been reading this paper here which mentions that they have used 20 MFCC coefficients from 40 mel ...
0
votes
1answer
54 views

How can you perform echo cancellation on features of audio signals?

I am working on a project that takes some features of audio signals and classify them with a neural network. However, the audio signals are added with their own echo with the environment. What would ...
0
votes
0answers
39 views

Average spectral shape

In this article about pre-emphasis on here, its use is defined as "to compensate for the average spectral shape". My question is, how is the average spectral shape calculated? This is the ...
0
votes
0answers
27 views

Matlab: separate parts of a speech signal vector automatically

I have a speech signal that contains 5 vowels consecutively. In order to do formant estimation of each vowel, I want to cut them out of the signal into a new variable. When looking at the plot of the ...
0
votes
1answer
64 views

Determining statistically the bandwidth of a signal from a spectrogram

My question is how to determine the parameters of a signal from a spectrogram statistically. The most important parameters that I am looking for are the lower and higher frequencies of the signal. I ...
0
votes
0answers
15 views

Source Filter Modelling of Speech Production

I am reading about mathematically modeling Speech using Source- Filter model. I want to know how can we prove that that the source filter model becomes an LTI system provided the configuration of ...
0
votes
0answers
26 views

Communicate to user that SNR is too low - Speech procesing

I am working in a project to estimate the signal to noise ratio (SNR) for an automatic speech recognition system in order to generate an alert to the user if the noise is too high in the place where ...
0
votes
0answers
35 views

Sinusoidal tone in Signal-to-Noise measurements

Mengapa kita menggunakan nada uji sinusoidal saat mengukur signal-to-noise pada saluran bicara daripada hanya sinyal ucapan itu sendiri ? EDIT: Translation from Indonesian according to Google: Why do ...
0
votes
0answers
25 views

How to identify the vocal folds in a mel spectrogram image?

I am quite new in the dsp and trying to learn How to identify the vocal folds in a mel spectrogram image? My image data looks like this. How to find the vocal folds in this image? I am using Praat ...
0
votes
0answers
54 views

How to use cross-correlation to verify speech data and determine delay

I have been experimenting with cross-correlation function to verify the presence of speech in a recorded file wrt a source file. I tried the following in Matlab: <...
0
votes
1answer
35 views

LPC Formant Estimation: Why resample?

The Burg algorithm implemented in PRAAT for formant estimation also works via LPC. First the sound is resampled to a sampling frequency of twice the value of maximum formant. Then a pre-emphasis is ...
0
votes
0answers
18 views

Help in LBG Vector Quantization - Splitting Factor

I'm currently making a program for speech recognition. In the step of codebook generation using the LBG (Linde-Buzo-Gray) algorithm, I've read that the splitting factor $\varepsilon = 0.01$ (generally)...
0
votes
1answer
35 views

How to use MFCC (or any other tool/method/algorithm) for speech verification

I have an original speech .wav file and a few recorded samples of the same file. Each recorded version has some differences wrt the original file. For example, The ...
0
votes
0answers
50 views

How do I extract n-dimensional cepstral coefficients with a sliding Hamming window in PyTorch?

I am trying to replicate the features from this paper, described as follows: MFCCs of speech frames are computed using a mel-scale filterbank. We extract 20-dimensional cepstral coefficients with a ...
0
votes
0answers
31 views

Compare two spoken words with MFCC and DTW

Processing of the speech signal is done using Matlab. I want to use MFCC (Mel-frequency cepstral coefficients) is implemented using HTK and DTW (dynamic time warping v2.1) algorithms to make a speech ...
0
votes
2answers
69 views

Conceptually confused by LPC for speech: Do we synthesize by the inverse filter (FIR)?

I've been following the process of using LPC to analyse the speech and then synthesize the speech by swapping out coefficients over time. I'm purely concerned with pitched vowels at this point. ...
0
votes
0answers
25 views

how to implement Bark filter bank on the power spectrum for PLP feature extraction?

So the implementations for PLP feature extraction I have seen in python and Matlab, are as follows: suppose the FFT features of an audio signal are of the size 257 (512 point FFT). if we have one ...
0
votes
0answers
79 views

STFT of ISTFT in the Griffin Lim Algorithm

The Griffin-Lim algorithm for phase recovery (based on the magnitude of an STFT) involves a step that is: STFT(Inverse STFT(...)). This seems to be the key iteration in the algorithm. This Quora ...
0
votes
0answers
29 views

Techniques for removing noise aside from Spectral Subtraction/Statistical Method?

I am studying how to remove noise from speech audio and I figured the most common error in spectral processing comes in the form of musical noise. Now what are the techniques on reducing this ...
0
votes
0answers
25 views

Capture the ambient sound

I sit in a cafe and start recording my voice, reading a poem. Then I sit at my room, at complete isolation, and I read the same exact poem once again and record it. Now, Considering I have these two ...
0
votes
0answers
19 views

Is audio normalisation required for a speaker recognition/speech classification task? How to do it?

Not entirely sure if this Stack or a ML-related stack would be a fit for this question, but I focus on the signal processing here on this question, so I decided this one I'm developing a speech ...
0
votes
0answers
33 views

Why do we use the term “cepstral” in MFCCs, and how does it related to the cepstral domain?

As I know, we can get the cepstrum by taking the Inverse Discrete Fourier Transform (IDFT) of the log of the power spectrum. And we can get the Mel-frequency cepstral coefficients (MFCCs) by taking ...