Generalized Residual Vector Quantization For Large Scale Data

Liu Shicong, Shao Junru, Lu Hongtao. Arxiv 2016

[Paper]    
ARXIV Quantisation Supervised Survey Paper

Vector quantization is an essential tool for tasks involving large scale data, for example, large scale similarity search, which is crucial for content-based information retrieval and analysis. In this paper, we propose a novel vector quantization framework that iteratively minimizes quantization error. First, we provide a detailed review on a relevant vector quantization method named \textit{residual vector quantization} (RVQ). Next, we propose \textit{generalized residual vector quantization} (GRVQ) to further improve over RVQ. Many vector quantization methods can be viewed as the special cases of our proposed framework. We evaluate GRVQ on several large scale benchmark datasets for large scale search, classification and object retrieval. We compared GRVQ with existing methods in detail. Extensive experiments demonstrate our GRVQ framework substantially outperforms existing methods in term of quantization accuracy and computation efficiency.

Similar Work