Approximately Nearest Neighborhood Image Search Using Unsupervised Hashing via Homogeneous Kernels

Table 1

HKLSH algorithm.

---

Process data

 (a) Obtain matrix from the database throughout the points

 (b) Obtain by randomly sampling a subset from

 (c) Project onto the th subset to obtain

 (d) Obtain

 (e) Project onto the homogeneous kernel space

 (f) Obtain the hash bucket

Query processing

 (a) Obtain the same hash bucket from the database

 (b) Use ANN (approximate nearest neighborhood) search for

     query matching

---