Detecting and mitigating audio-visual incongruence

A non-transitory computer-readable medium storing instructions that, upon execution by one or more processors, causes the one or more processors to perform operations of at least partially ambiguating spatial audio cues in audio in response to determining that incongruence may be present, the operations comprising:

• capturing, by a camera of a video conferencing system, video images of a conference room;
• capturing, by an audio device, an audio stream from the conference room;
• determining, by a processor of the video conferencing system, that potential incongruence is present in far end play back of the captured video images together with playback of the audio stream, the potential incongruence being determined based on a location of the audio device within the conference room;
• and
• filtering, by the processor, the captured audio stream in response to the determining that potential incongruence is present, the filtering comprising:
• decoding the captured audio stream into a plurality of channels, the plurality of channels having a structured spatial relationship;
• assigning random phase values that are band-limited to each of the plurality of channels to modify the captured audio stream;
• constructing filters by interpolating a full filter response between the assigned random phase values;
• and
• applying the constructed filters to the captured audio stream to provide an ambiguated audio stream corresponding to the captured audio stream.

A video conferencing system configured to at least partially ambiguated spatial audio cues in audio in response to determining that incongruence may be present, the video conferencing system comprising:

• a camera configured to capture video images of a conference room;
• an audio device configured to capture an audio stream from the conference room;
• and
• a processor configured to perform operations comprising:
• determining that potential incongruence is present in far end play back of the captured video images together with playback of the audio stream, the potential incongruence being determined based on a location of the audio device within the conference room;
• and
• filtering the captured audio stream in response to the determining that potential incongruence is present, the filtering comprising:
• decoding the captured audio stream into a plurality of channels, the plurality of channels having a structured spatial relationship;
• assigning random phase values that are band-limited to each of the plurality of channels to modify the captured audio stream;
• constructing filters by interpolating a full filter response between the assigned random phase values;
• and
• applying the constructed filters to the captured audio stream to provide an ambiguated audio stream corresponding to the captured audio stream.