Locality-sensitive hashing scheme based on dynamic collision counting 论文
2012引用 226
Advanced Image and Video Retrieval TechniquesAlgorithms and Data CompressionData Management and Algorithms
摘要
Locality-Sensitive Hashing (LSH) and its variants are well-known methods for solving the c-approximate NN Search problem in high-dimensional space. Traditionally, several LSH functions are concatenated to form a "static" compound hash function for building a hash table. In this paper, we propose to use a base of m single LSH functions to construct "dynamic" compound hash functions, and define a new LSH scheme called Collision Counting LSH (C2LSH). If the number of LSH functions under which a data object o collides with a query object q is greater than a pre-specified collision threhold l, then o can be regarded as a good candidate of c-approximate NN of q. This is the basic idea of C2LSH.