Anti-sparse coding for approximate nearest neighbor search

TL;DR

This paper introduces an anti-sparse coding-based binarization method for high-dimensional vectors, enabling explicit reconstruction and outperforming traditional hashing techniques in approximate nearest neighbor search.

Contribution

It presents a novel binarization scheme using anti-sparse coding that allows explicit reconstruction and demonstrates superior performance over random projections in high-precision scenarios.

Findings

Anti-sparse coding enables explicit vector reconstruction from binary form.

The proposed method outperforms Locality Sensitive Hashing with regular frames.

High-dimensional binarization benefits from the new scheme in accuracy.

Abstract

This paper proposes a binarization scheme for vectors of high dimension based on the recent concept of anti-sparse coding, and shows its excellent performance for approximate nearest neighbor search. Unlike other binarization schemes, this framework allows, up to a scaling factor, the explicit reconstruction from the binary representation of the original vector. The paper also shows that random projections which are used in Locality Sensitive Hashing algorithms, are significantly outperformed by regular frames for both synthetic and real data if the number of bits exceeds the vector dimensionality, i.e., when high precision is required.