Time delay neural network for printed and cursive handwritten character recognition.

I. Guyon, J. S. Denker, and Y. Le Cun.
US Patent 5,105,468.

A time delay neural network is defined having feature detection layers which are constrainted for extracting features and subsampling a sequence of feature vectors input to a particular feature detection layer. Output from the network for both digit and uppercase letters is provided by an output classification layer which is fully connected to the final feature detection layer. Each feature vector relates to coordinate information about the original character preserved in temporal order together with additional information related to the original character at the particular coordinate point. Such additional information may include local geometric information, local pen information, and phantom stroke coordinate information relating to connecting segments between the end point of one stroke and the beginning of another stroke.

The network is also defined to increase the number of feature elements in each feature vector from one feature detection layer to the next. That is, as the network is reducing its dependence on temporally related features, it is increasing it dependence on more features and more complex features.

7 Claims, 7 Drawing Sheets.

[ next paper ]