Audio watermark detection for delivering contextual content to a user

A system for delivering content to a mobile communication device comprising software for detecting audio watermark in a watermarked audio signal broadcasted by a local machine and if necessary triggering an event on the mobile communication device, the system having software, that when executed by a computing environment, being adapted to comprise:

• the mobile communication device, periodically determining, a user local profile, the local profile comprising a localization of the mobile communication device;
• the mobile communication device periodically sending, to a remote server, a request for parameters of at least one audio watermark i, the parameters comprising N frequency bands Fik, k=0, . . . , N−1, a likelihood function Li to be computed, a likelihood threshold Ti and at least one event to be triggered on the mobile communication device corresponding to the audio watermark along with the user local profile;
• the mobile communications device receiving the parameters form the remote server, when necessary, if necessary, the parameters having been selected by the remote server in view of the user local profile so that the event is relevant for the user;
• the mobile communication device storing in its memory the parameters;
• the mobile communication device polling its environment for acquiring an audio signal and processes the audio signal in:
• determining the power spectral density of the audio signal for each of the frequency bands;
• computing the likelihood function corresponding to each watermark to be studied;
• if the result of the likelihood function exceeds the likelihood threshold, triggers the corresponding event on the mobile communication device.