Unsupervised Learning of Temporal Features

It is easy to recognize a radio channel that carries Morse code, even if you do not know the code. Morse code consists of a small set of simple temporal features (a dot, a dash, and two pause lengths), and a Morse signal is built by their repeated occurrences. The brain quickly identifies these features as the dominant content of the received signal. We have implemented an optical system that can discover, on its own, the dominant features in a temporal signal characterized by repetitive entities. This task is a precursor to the more complex processing required for the self-organized feature extraction and recognition of audio and sonar signals.

Become a member or log in to view the full text of this article.

OSA Members get the full text of Optics & Photonics News, plus a variety of other member benefits.

Publish Date:

Unsupervised Learning of Temporal Features

It is easy to recognize a radio channel that carries Morse code, even if you do not know the code. Morse code consists of a small set of simple temporal features (a dot, a dash, and two pause lengths), and a Morse signal is built by their repeated occurrences. The brain quickly identifies these features as the dominant content of the received signal. We have implemented an optical system that can discover, on its own, the dominant features in a temporal signal characterized by repetitive entities. This task is a precursor to the more complex processing required for the self-organized feature extraction and recognition of audio and sonar signals.

This article is only available as a PDF.

Download PDF

Publish Date:


Add a Comment

ADVERTISEMENT
ADVERTISEMENT