Abstract:
A method of generating an audio visual identification comprising the steps of extracting an audio sequence from a randomly generated noise generator, such as a white or pink noise generator, and saving the sequence as an audio tone; attaching the audio tone to a data file, which data file has a binary value; allocating a binary value to the combined audio tone and data file, whereafter the combined binary value enables the verification of the data file content at time of audio tone attachment. Where data file modification, alternately a different audio tone is attached to the data file – the binary value would not correspond to the binary value of the initial combined audio tone with data file. Data verification is thereby confirmed where binary values correspond.
Abstract:
A recording device, such as a video camera, emits audio watermarks (e.g., audible tones, inaudible tones) and records audio watermarks emitted by other recording devices. The audio watermarks provide information (e.g., data) such as an identifier of the camera and the identity of the operate. Audio watermarks may further include time information such as a timestamp indicating the time that the audio watermark was emitted by the camera. Audio watermarks may be emitted and recorded at intervals while recording an incident. A processing system, such as a digital video and audio processing system, may receive video information and recorded watermarks from multiple recording devices. The processing system may use the audio watermarks recorded by each recording device to determine which recording devices were in the vicinity of an incident. The processing system may also use the audio watermarks to align (e.g., synchronize) the video recorded by the various recording devices. Alignment of recorded video may permit synchronized playback of the recorded video from the various recording devices. Alignment may further identify at what point during an incident a recording device began and terminated recording.
Abstract:
Methods and devices are described for losslessly watermarking an audio signal by performing a noise shaped quantisation and clipping the output from the noise shaped quantisation to bounds computed by a pair of quantised linear functions with gradient 0.5 of the input to the noise shaped quantisation. Corresponding methods and devices are also described for inverting the process to recover an exact replica of the original audio signal.
Abstract:
An enterprise voice system such as a contact centre is disclosed which provides a speech analytics capability. Whilst call recording is common in many contact centres, calls are normally recorded in single-channel audio files in order to save costs. Previous attempts to provide automatic diarization of those recorded calls have relied on training the system to recognise voiceprints of users of the system, and then comparing utterances within the recorded calls to those voiceprints in order to identify who was speaking at that time. In order to avoid the need to train the system to recognise voiceprints, an enterprise voice system is disclosed which inserts a digital watermark into the digitised audio signal from each user's microphone. By inserting the digital watermark with an energy, and, in some cases also with a spectrum, which matches the digitised audio signal, and taking advantage of typically only one user speaking at a time, a mark is left in the recorded call which a speech analytics system can use in order to identify who was speaking at different times in the conversation.
Abstract:
The present invention relates to a method of watermarking an audio track by providing a database structure, assigning a unique number X to the audio track and partitioning the unique number X into N positions i, and N values v,, where i=1...N. For each position i of the number X, the method comprises generating a pseudorandom number R(i, v i ) based on the position i and the value v, at the position i, adding a new node [v i ,R(i,V i )] at level i in the database structure, if not already present in the database structure, adding an edge from the preceding node [V i-1 , R(i-1,V i-1 )] to the new node [v i ,R(i,V i )], generating, by using R(i, v i ,) as a seed, a set of random complex numbers C, inserting the set into a vector V, calculating the inverse Fourier transform of vector V, and adding the thereby obtained time domain function y(t) to the audio track.
Abstract:
Playback of an audio stream is synchronized on multiple connected digital devices by using synchronization fingerprints. Playback actions may furthermore be synchronized on all devices, such as skips and pauses. Furthermore, synchronization may be maintained even in the presence of variations in decoding speed, playback interruptions, and network disconnections. Synchronized playback of streamed audio content on multiple devices is achieved by devices compensating for time drifting induced by network instability and variable playback speed across master and guest devices to reduce the formation of echoes during playback.
Abstract:
A method of communicating data imperceptibly in an audio signal. The method comprises, for each sub-band of the audio signal, identifying the tone in that sub-band having the highest amplitude. An audio code comprising the data to be communicated is scaled by a frequency mask profile, the frequency mask profile having maxima at the frequencies of the identified tones. The audio signal and the scaled audio code are aggregated to form a composite audio signal. The composite audio signal is then transmitted.
Abstract:
There is disclosed a method of establishing a communication link between first and second devices, the method comprising: receiving a link request from a first device with a first unique identifier; allocating a media content from multiple media content to the first unique identifier; sending the media content to the first device; receiving a media identifier from a second device with a second unique identifier; comparing the media identifier and the media content, and in dependence on identifying a match, transmitting the first unique identifier to the second device and the second unique identifier to the first device, whereby the first and second device can establish a communication link based on the exchanged first and second unique identifiers.