Audio fingerprint matching query method and device

Abstract

The invention discloses an audio fingerprint matching query method. The audio fingerprint matching query method comprises the following steps: counting query times of each kind of audio information within a set counting period, and sequentially ranking the audio information from high to low according to the query times; building a hot audio model for top n audio information, wherein n is a positive integer; loading the hot audio model into a hot audio server; when receiving a query request transmitted by a client, extracting an audio fingerprint in the query request, and searching the audio information matched with the audio fingerprint from the hot audio model in the hot audio server; and returning the matched audio information to the client. The invention further discloses an audio fingerprint matching query device. The hot audio model is built in the system, so that reduction of the modeling pressure of the system is facilitated. The audio information is searched by the hot audio server from the hot audio model, so that improvement of the throughput of the system is facilitated; and the audio matching query efficiency and success rate are improved.

Description

Technical field [0001] The present invention relates to the field of Internet technology, in particular to a method and device for audio fingerprint matching and query. Background technique [0002] With the development of the Internet and multimedia technology, users can use audio players, browsers and other clients to search for music they need on the Internet. The search method is usually to enter information such as the name of the music, the name of the artist, and the album name. At present, due to the development of audio fingerprint technology, a new type of audio retrieval method has emerged. When the user plays a piece of audio on the audio player, browser and other clients, if the source of the audio is not from the audio library in the corresponding server of the client, the information corresponding to the audio will not be displayed directly on the client. The client can extract the audio fingerprint from the audio being played and send it to the server for matchin...

AI Technical Summary

Problems solved by technology

In order to achieve efficient matching and query capabilities, the most desirable situation is to deploy a model on one server, that is, 18 servers are required, resulting in increased resource costs

In order to save server resources, one server usually deploys multiple models at present, which increases the workload of the server when matching queries, reduces system throughput, and affects the efficiency and success rate of audio matching queries

Images