Collaborative Develpment of Speech and Language Technology

This website is a multi-national, multi-lingual collaboration of a number of groups of students, entrepreneurs, early stage companies, and interested individuals. It focuses on human and machine learning about speech and language in many languages. The human learning is based on the principle of learn-by-doing. Individually and in small teams, the students will complete a number of projects of substance with real-world value up to and including entrepreneural projects producing an MVP (minimal viable product) and academic projects publishable in peer-reviewed journals. The founding principle of this collaboration is the principle of cooperation as embodied in the philosophy of the Golden Rule and the allegory of the Long Spoons.

Mathematically, these principles are represented by the theory of cooperative n-person games of von Neumann and Morgenstern as opposed to the Nash Equilibrium theory. With this philosophy, this collaboration hopes to grow by way of a virtuous pyramid in which the greatest rewards are at the base, rather than at the top, of the pyramid. If this growth is successful, this effort could eventually lead to speech and language technology being developed for hundres or even thousands of languages. The planned machine learning methodology will be a hybrid of hidden stochastic process modeling and deep learning based on deep neural networks (DNNs). This architecture is capable of implementation on a massively parallel, loosely connected network.



A Spectrogram Identification Quiz

Two spectrograms are shown below in a random order. A new random question is generated each time you press "Get New Question". If you don't know, press "Get Answer" and the correct answer will be given. The two words in each pair are acoustically confusable. They differ from each other in only one distinctive feature. Throughout the website, there will be many more examples of minimal pair words. Understanding these distinctions is at the core of understanding speech and how computers can learn to recognize or synthesize speech.

Which spectrogram is which of the following two words?

Spectrograms will go here

What is the first word? (The second is the other word.)

First word:
Your name:

Don't know the answer? Press the button, but then you can't score:

Your score will appear here.

Want to try another and improve your score?



Why take a spectrogram quiz?

Why study pairs of confusible words?