Audio Classification and Music Recognition

Audio classification and music recognition system enables automated monitoring of radio and TV broadcasts, ensuring compliance with media regulations and supporting large-scale content analysis through AI-powered processing.

Let's start a project

Services

Product DiscoveryProduct DesignTechnology ArchitectureSoftware DevelopmentAI DevelopmentQA & SecurityOperations & Support

Audio Classification and Music Recognition

Client

A large-scale organization operating in the media and communications sector, responsible for monitoring broadcast content and supporting the consistent application of regulatory and compliance requirements. The organization works with extensive volumes of media data and focuses on ensuring reliable oversight of radio and other broadcast channels, as well as enabling structured analysis of media content across long time spans.

Challenge

A significant portion of radio airtime must be filled with Hungarian music: at least 35% on linear channels and 50% on community radios. In addition, a quarter of the music played annually must consist of works or recordings released in the past five years. Analyzing several years’ worth of available content would also make previously impossible media research feasible.

The existing IT infrastructure did not provide adequate tools for the fast and efficient monitoring of these requirements or for making media assets searchable. According to the project requirements, the system must be capable of processing the daily airtime of up to 1,000 channels within 24 hours.

Solution

A system was created that, after processing an external reference music database, is capable of recognizing music played on registered channels. Based on this, the system automatically generates a broadcast log, which can be exported in multiple formats. This has enabled further analysis and research opportunities.

The system runs 24/7, distributing the processing of configured channels across multiple machines if needed.

The system’s functions and settings are accessible and manageable via a web-based user interface.

Results

The system’s automated monitoring capability has been proven, significantly reducing the burden of manual work and making the monitoring process more efficient. With the live rollout of the system, tracking broadcasters’ airtime has become faster and more accurate.

Get in touch

Leave your contact details and we will reach out to you.