Vector quantization stands as an algorithmic approach in contrast to scalar quantization. In scalar quantization, the input is a single numerical value, and each quantizer codeword represents a solitary sample of the source output. Within the lossy compression framework, encoding a sequence of samples enhances the efficiency of the encoding process. This concept is also employed in quantization (as a lossy compression technique). Consequently, in vector quantization, a sequence of samples is quantized using a codeword of length L, rather than using individual samples. In this project, we have employed K-means clustering to implement vector quantization.

Keywords: Vector Quantization; Scalar Quantization; Lossy Compression; Linde-Buzo-Gray
Course: Data Compression
Code

Figures and Tables

• Method:

• Result: