11
votes
Accepted
What's the correct graphical interpretation of a series of MFCC vectors?
The way MFCC's are always used is by feeding them into the classifier. This can be done on a frame-by-frame basis (12x1 vector), or by concatenating (...
- 10.5k
9
votes
Accepted
Why is each window/frame overlapping?
Why is each window/frame overlapping?
Windowing is a means to stationarize signals. Inside a small enough window, you can expect that the properties of the signal chunk do not vary too fast. And you ...
- 30.3k
9
votes
What are i-vectors and x-vectors in the context of Speech Recognition?
The i-vectors and x-vectors share the ability to represent speech utterance in a compact way (as a vector of fixed size, regardless of length of the utterance).
The extraction algorithms of i-vectors ...
- 231
7
votes
How to do continuous signal processing (i.e without windowing)?
Yes, it's possible to analyse sound the way ears do.
For example, you could compute the DFT of a signal continuously using several Goertzel filters.
$$
y_k[n] = e^{j2\pi k/N} y_k[n-1] + x[n]
$$
...
- 22.9k
7
votes
Why is each window/frame overlapping?
More overlap means you end up with more windows (of a given length) per second of audio. More windows (of a given length) requires more FFTs which requires more MACs or FLOPs which generally requires ...
- 34k
6
votes
VAD speech databases
Researchers from the Johns Hopkins University have recently released a corpus of music, speech, and noise which, according to them, is suitable for training models for voice activity detection and ...
- 61
6
votes
How does Siri recognize me saying "Hey Siri"?
"Ok Google" is described in many publications by Google
Automatic Gain Control and Multi-style Training for Robust Small-Footprint Keyword Spotting with Deep Neural Networks
Convolutional Neural ...
- 1,119
6
votes
Accepted
What Is the Point of Doing the Zero Padding?
You may think of it as efficient way to apply Dirichlet Window Based interpolation in the Fourier Domain.
The advantage of applying the interpolation using Zero Padding in the Time Domain is very ...
- 41.4k
5
votes
Accepted
Using HMM for speech recognition
What exactly can the vector/observation be? Can it be a relatively simple function, say approximation of the spectral envelope? Or is it usually a vector of R^n?
You won't go very far by summarizing ...
- 19.1k
4
votes
Accepted
VAD speech databases
You can take some big speech corpus like TEDLIUM and add the noise you like:
http://www-lium.univ-lemans.fr/en/content/ted-lium-corpus
The advantage of TEDLIUM is that it's a set of continuous ...
- 1,119
4
votes
Accepted
What is best practice to remove static noise?
The standard method to remove stationary noise is spectral subtraction, where
the magnitudes of the short-time Fourier transform of the noisy signal are modified based on an SNR estimate in the ...
- 80.4k
4
votes
Accepted
In framing of audio samples', what is need of frame shift while giving frame size??
By performing the windowing with overlap we are artificially increasing our time resolution (larger granularity of features in time). This is especially useful when frame duration is long (bad time ...
- 10.5k
4
votes
Accepted
Averaging Different Length Speech Signal
Dynamic Time Warping is pretty well explained on this site. I'll use some of the diagrams from the PPT on that site to explain.
The idea is to divide the signals into segments (frames) and then ...
- 22.9k
4
votes
Cepstrum calculus disambiguation
A common technique for computing the inverse fft is to invert the imaginary part of the input array, perform a forward fft and then invert the imaginary part of the output array. In the case of the ...
- 486
3
votes
Accepted
Initial state recognition in HMM
The Baum-Welch algorithm uses the EM (Expectation Maximization) algorithm to estimate the model parameters $(T, E, \pi)$, where:
$T$: the transition probabilities
$E$: the emition probabilities
$\...
- 236
3
votes
Remove a known wav file from recorded file
Assuming that there are no non-linear effects between the sound N source and the microphone (such as AGC), you might try to estimate the impulse response of the channel between the source for sound N ...
- 34k
3
votes
Accepted
How to edit audios so that they have the same length
ImageMagick is a great tool to edit multiple images in the command line. Thanks to your question, I just discovered SoX, or Sound eXchange:
SoX is a cross-platform (Windows, Linux, MacOS X, etc.) ...
- 30.3k
3
votes
Is there any way that we can perform speech recognition without using Fourier transforms?
All recognition tasks (doesn't even have to be speech recognition) are reductions of a very high-dimensional signal (your speech recording's dimension is the number of audio samples!) to a low-...
- 26.6k
3
votes
Accepted
filterbank: understand the different responses at the center frequency of each filter
Normalizing the filterbanks by their widths is optional and totally up to you (similarly to the warping scale Mel/Bark). Depending on your application, you can start without normalization and see what ...
- 10.5k
2
votes
What do cloud-based speech-to-text services upload to the cloud?
To give more variants to Aaron's answer, speech recognition pipeline has multiple stages where you can cut the line between client and server. There are the following variants:
Lossless audio (flac)
...
- 1,119
2
votes
What do cloud-based speech-to-text services upload to the cloud?
Its probably safe to assume that all of the major companies send enough information to reconstruct the audio. This is because having that audio for training is such a valuable resource. A certain ...
- 1,382
2
votes
Call recognition using Digital Signal Processing
See the article below for one of the methods we use for detecting andn classifying dolphin whistles. These methods may have some sucess with birds, but I can promise that they won't separate 300 ...
- 131
2
votes
Accepted
The logarithm Energy measure in Speech
You compute the energy (before the log operation) as
$$\max\left\{\sum_i s^2(i),\epsilon\right\}\tag{1}$$
with some small constant $\epsilon$ (they suggest $\epsilon=2e-22$ which gives $\ln\epsilon\...
- 80.4k
2
votes
Speech recognition using mfcc and neural network
Usually your training data and your test data are different, otherwise your results will be unrealistically good. However, for a first test of your implementation, you can simply use your training ...
- 80.4k
2
votes
Accepted
Filtering noise and identifying phones boundary in speech recongition
system using arduino
Arduino performance is not enough for any serious speech recognition. You should better play with your mobile phone instead. With mobile you can build at least some interesting ...
- 1,119
2
votes
VAD speech databases
Free VAD Data set recorded in real environment with ground truth label:
refer: https://github.com/jtkim-kaist/VAD-Toolkit
- 31
2
votes
Accepted
Continuous speech recognition with HMM
Will similar enough vectors be "put together" and only treated as 1?
No, this is not how it is done.
so for what p will P(o|p) be computed?
Your assumption that recognition is performed by ...
- 19.1k
2
votes
Vowel detection using MFCC
You may just implements a VAD (voice activity detector) whose parameters are high enough to filter out voiced consonants. One VAD which always amazed me is the zero crossing rate, because it is so ...
- 61
2
votes
Accepted
How to emulate human sound recognition
I do not think that there is a standard approach to this problem, because as far as I know it has not yet been solved satisfactorily. I believe that you can find a solution to a (very) simplified ...
- 80.4k
Only top scored, non community-wiki answers of a minimum length are eligible