In this paper, designs and analyses of compressive recognition systems arediscussed, and also a method of establishing a dual connection between designsof good communication codes and designs of recognition systems is presented.Pattern recognition systems based on compressed patterns and compressed sensormeasurements can be designed using low-density matrices. We examine truncationencoding where a subset of the patterns and measurements are stored perfectlywhile the rest is discarded. We also examine the use of LDPC parity checkmatrices for compressing measurements and patterns. We show how more generalensembles of good linear codes can be used as the basis for pattern recognitionsystem design, yielding system design strategies for more general noise models.
展开▼
