26

Yes, there is some research on this topic, which can be called adversarial machine learning, which is more an experimental field. An adversarial example is an input similar to the ones used to train the model, but that leads the model to produce an unexpected outcome. For example, consider an artificial neural network (ANN) trained to distinguish between ...


12

Sometimes if the rules used by an AI to identify characters are discovered, and if the rules used by a human being to identify the same characters are different, it is possible to design characters that are recognized by a human being but not recognized by an AI. However, if the human being and AI both use the same rules, they will recognize the same ...


10

Yes there are, for instance one pixel attacks described in Su, J.; Vargas, D.V.; Kouichi, S. One pixel attack for fooling deep neural networks. arXiv:1710.08864 One pixels attacks are attacks in which changing one pixel in input image can strongly affect the results.


5

Here's an example: How to hack your face to dodge the rise of facial recognition tech In his recent book The Fall, Stephenson wrote about smartglasses that that project a pattern over the facial features to foil recognition algorithms (which seems not only feasible but likely;) Here's an article from our sponsors, Adversarial AI: As New Attack Vector ...


4

Your problem is an old one. There are many methods, referred to as Voice Activity Detection (VAD) methods, which detect speech from an audio signal. The typical design of a VAD algorithm follows one of more of these 3 approaches: A noise reduction stage, e.g. via spectral subtraction. Some features or quantities are calculated from a section of the input ...


4

In machine learning you normaly split your data into 3 parts(80-10-10%). First part is for the training of your ML-model. The second part (10%) is the development set (or validation set). This is used as measuring your performance with various hyper parameters (e.g. in neural networks: layer size). After you found your best hyper parameters, you learn the ...


4

There are many insightful comments and answers so far. I want to illustrate my idea of "color blindness test" more. Maybe it's a hint to lead us to the truth. Imagine there are two people here. One is colorblind (AI) and another one is non-colorblind (human). If we show them a normal number "6", both of them can easily recognize it as number 6. Now, if we ...


4

Isn't that essentially what chess does? For example, A human can recognize that a Ruy exchange offers white great winning chances (because of pawn structure) by move 4 while an engine would take several hours of brute force calculation to understand the same idea.


3

Here's a live demo: https://www.labsix.org/physical-objects-that-fool-neural-nets/ Recall that neural nets are trained by feeding in the training data, evaluating the net, and using the error between the observed and the intended output to adjust the weights and bring the observed output closer to the intended. Most attacks have been on the observation that ...


3

Fourier transform is used to transform audio data to get more information (features). For example, raw audio data usually represented by a one-dimensional array, x[n], which has a length n (number of samples). x[i] is an amplitude value of the i-th sample point. Using the Fourier transform, your audio data will be represented as a two-dimensional array. ...


3

Yes it can. However, other differences between training and test data with audio could have greater effect: Identity of the speaker (including effects from gender, age, physical build, local accent, amongst others) Acoustics of the recording environment (including proximity to the microphone, size of space, presence of hard surfaces, background noise) If ...


2

Yes, there is. An extremely quick search found this: Multimodal Speaker Identification Based on Text_and_Speech. Let me tl;dr for you: (My abstract addition in Italics) Novel method for speaker identification based on both speech utterances and their transcribed text. They first transcribed text of each speaker’s is processed by using probabilistic ...


2

I can't speak to wit.ai specifically, but I can tell you a little bit about how similar applications work. Specifically, I can talk a bit about Apache Stanbol which also converts free text into structured data. That said, I should prefix this by saying there isn't just one way to "get there from here." Many techniques could be part of a stack for ...


2

Yes, it is possible, even if the best approach could be different from neural networks. Anyway, you should extract some significant features from the audio (energy, onsets, root frequencies, and other). Usually, more features than those really needed are extracted and afterwards the most sigificant are selected through some algorithm (e.g. PCA). In this way ...


2

I would advise you to look into Mozilla’s implementation of Baidu DeepSpeech here


2

According to what Josh Dotson posted via medium,gives a clear insightful knowledge concerning the following; 1.Speech data besides speech recognition. Language modelling. Text to speech. Machine translation. Signal processing. And lastly, books and blogs for further research Resources for acknowledgement


2

I don't know about voice recognition but for NLP i think that Gensim could be what you are looking for! Gensim is a NLP package that contains efficient implementations of many well known functionalities for the tasks of topic modeling such as tf–idf, Latent Dirichlet allocation, Latent semantic analysis... About the readings, maybe you can start with the ...


2

Summarizing text is always going to be 'easier or more efficient' than voice simply because voice requires the additional step of converting to text. That doesn't tell you anything about accuracy. From an article published on June 1, 2017, Google’s speech recognition is now almost as accurate as humans: "According to Mary Meeker’s annual Internet Trends ...


2

If you have fixed length speech data you can detect the content using only CNN. You can see that problem as a binary classification (1 if the spoken word is correct, 0 otherwise). But first, you need to make the input length is fixed. For example, you use 2 seconds as the fixed length. If the recorded speech is more than 2 seconds, you need to crop it, and ...


2

There are some research at least on the "foolability" of neural networks, that gives insight on potential high risk of neural nets even when they "seem" 99.99% acurate. A very good paper on this is in Nature: https://www.nature.com/articles/d41586-019-03013-5 In a nutshell: It shows diverse exemples of fooling neural networks/AIs, for exemple one where a ...


1

The reason for robot like speech may be because tacotron uses griffin lim for vocoder, which cannot reproduce sound with perfection, often introducing robot like sound artifects. A vocoder is a network that transforms a transform a spectrogram image back to speech waveform. Tacotron and many other speech generation neural network uses CNN to generate ...


1

The tutorials you link are not much relevant, there are already existing implementations of your exact problem. You can use https://github.com/swshon/dialectID_e2e, there are many other similar implementations on github.


1

In the field of Automatic Speech Recognition (ASR) Kaldi is the current leader. Before Deep Neural Network era there were Sphinx and HTK.


1

The most usual differences in signal records caused by different microphones will have small if not null impact in the recognition accuracy, in particular if we are talking about changes one mic by another of same model and manufactor: Differences in bandwidth: voice is in a very common (central) bandwidth, it is not expected these differences impacts, even ...


1

your question is a very similar to "turing-test". you could narrate a simple story and ask questions based on that , considering the state-of-art algorithms in "question-answering" are still far beyond human skills.


1

The difference between a Speech API and a Speech Engine is: Speech API's enable developers to integrate speech recognition technologies into developer apps. On the other hand a speech engine is software that gives your computer the ability to play back text in a spoken voice. (Source msdn library) Below is a list of speech recognition tool-kits and their ...


1

There are probably many different approaches to doing what you are talking about, but the most common would probably be using long short term memory recurrent neural networks to operate on the data through time. You would be able to train your neural network to do the things that you want to do such as classify the speaker, and classify noises into different ...


1

Edit: It's not clear what exactly you're trying to accomplish... My answer assumes you wanted to split the man and the woman's audio, but re-reading your question make me think otherwise. Note: Don't expect this to be a perfect answer. I'm not an expert in the field, just an interested student. I can't comment on this site yet, so I'm submitting this as ...


1

Speaker identification is quite widely researched domain. Modern approach would be to map speaker information to i-vector, a real-valued vector of 200-400 components that characterizes speaker fully. i-vectors allow very precise speaker identification and verification. For more information you can check i-vector tutorial Also you can check state of the art ...


Only top voted, non community-wiki answers of a minimum length are eligible