Vector Quantization is a lossy compression scheme that represents data as a list of elements from a set of vectors. For example, an image might be represented as indices into a dictionary of 2x2 pixel blocks. Advantages of this scheme include relatively faithful reproduction, and fast decompression. Disadvantages include poor reproduction for certain documents (this is true of all lossy compression schemes) and slow compression. The algorithm presented in the optimal quantisation node works for vector quantization as well; just substitute the word vector for the word color.

Log in or register to write something here or to contact authors.