A Tunable, Efficient, Specialized Multidimensional Range Query Algorithm

Abstract

This paper presents algorithms for fast indexing and searching of a metadata database using a particular type of audio fingerprint as the key. Because of the variability of audio and hence its content-derived fingerprints, simple index-based database lookups have previously been proven to be insufficient. Prior approaches, on the other hand, are computationally complex and rely on large memory and processing resources for realistic applications, which makes them unsuitable for deployment in embedded environments. Thus, a multidimensional range query algorithm has been developed to efficiently cope with altered audio content. This search algorithm is tunable to adjust desired coverage, CPU load, and disk load. Special care has been taken to design the system to be suitably fast and lightweight to implement on extremely modest computing hardware, such as portable consumer electronic devices.

Recognition of Activities of Daily Living Based on Environmental Analyses Using Audio Fingerprinting Techniques: A Systematic Review

“…In [ 48 ], the authors proposed an algorithm with two stages named indexing and search. The indexing is based in the construction of zone tables using the Search by Range Reduction (SRR) threshold values [ 48 ].…”

“…In [ 48 ], the authors proposed an algorithm with two stages named indexing and search. The indexing is based in the construction of zone tables using the Search by Range Reduction (SRR) threshold values [ 48 ]. The search is based on the SRR test, calculating the Itakura distance between two fingerprints, and comparing it with values in the zone tables [ 48 ].…”