Searching with quantization: approximate nearest neighbor search using short codes and distance estimators

We propose an approximate nearest neighbor search method based on quantization. It uses, in particular, product quantizer to produce short codes and corresponding distance estimators approximating the Euclidean distance between the orginal vectors. The method is advantageously used in an asym- metric manner, by computing the distance between a vector and code, unlike competing techniques such as spectral hashing that only compare codes. Our approach approximates the Euclidean distance based on memory e- cient codes and, thus, permits ecient nearest neighbor search. Experiments performed on SIFT and GIST image descriptors show excellent search accuracy. The method is shown to outperform two state-of-the-art approaches of the lit- erature. Timings measured when searching a vector set of 2 billion vectors are shown to be excellent given the high accuracy of the method.