Invariance of WNNs Performance for Pattern Recognition Testing Translated Image Input Spaces

摘要

Weightless Neural Networks (WNNs) have been used in pattern recognition vision systems for many years. In operation, a series of images of an object are presented to the network, each being processed suitably and effectively stored in a memory called discriminator. Then, when another image is shown to the system, it is processed in a similar manner and the system reports how similar it is to the taught images The operation of these networks requires that binary values be produced from the input data (processing data). In this paper, the network generalisation properties,that is recognising unseen images, are studied for translated images.