Overview
The Audio Content Recognition System performs automatic recognition of audio content from analog or digital file sources. Application examples include advertisement tracking and recording, music (song) recognition, and copyright monitoring. Content sources can be analog, for example radio or television inputs, digital, for example network, telephony, or high-speed serial connections, or file based formats such as WAV, MP3, etc.
The system is based on a high-performance DSP board installed in a PC; the DSP board serves as recognition engine and the PC provides a "host" GUI for operating the system and remote communication with other systems. If analog sources are used, then the DSP board is fitted with an audio input module that allows direct A/D conversion and data feed to the 'DSP farm' on the board, which allows efficient processing for multichannel systems.
Real-time operation is supported for up to 8 channels per DSP board when database size is limited. Non-real-time operation is supported with larger databases.