Affinity Preserving Quantization for Hashing: A Vector Quantization Approach to Learning Compact Binary Codes | Awesome Learning to Hash Add your paper to Learning2Hash

Awesome Learning to Hash

A webpage dedicated to the latest research on learning-to-hash, including state-of-the-art deep hashing models, all updated on a weekly basis. Maintained by Sean Moran.

Contact Sean Moran about this survey or website. Made with Jekyll and Hyde.

Affinity Preserving Quantization for Hashing: A Vector Quantization Approach to Learning Compact Binary Codes

Z. Wang, L. Duan, T. Huang, W. Gao. AAAI 2016

[PDF]    
AAAI Quantisation

Hashing techniques are powerful for approximate nearest neighbour (ANN) search. Existing quantization methods in hashing are all focused on scalar quantization (SQ) which is inferior in utilizing the inherent data distribution. In this paper, we propose a novel vector quantization (VQ) method named affinity preserving quantization (APQ) to improve the quantization quality of projection values, which has significantly boosted the performance of state-of-the-art hashing techniques. In particular, our method incorporates the neighbourhood structure in the pre- and post-projection data space into vector quantization. APQ minimizes the quantization errors of projection values as well as the loss of affinity property of original space. An effective algorithm has been proposed to solve the joint optimization problem in APQ, and the extension to larger binary codes has been resolved by applying product quantization to APQ. Extensive experiments have shown that APQ consistently outperforms the state-of-the-art quantization methods, and has significantly improved the performance of various hashing techniques.

Similar Work