vector quantization

Vector quantization is a technique for approximating a large set of vectors by a smaller set of representative vectors (centroids). In AI, this is commonly employed in signal processing and neural network compression to reduce model size and improve efficiency.

5 papers