Methods and apparatus to detect spillover in an audience monitoring system
View Patent ↗Methods and apparatus to detect spillover in an audience monitoring system are disclosed. An example apparatus includes an audio sample selector to select first audio samples of a first audio signal received by a first microphone, the first audio signal associated with media; an offset selector to select first offset samples corresponding to a second audio signal received by a second microphone, the second audio signal associated with media, the first offset samples offset from the first audio samples by a first offset value; a cluster analyzer to, if a first count of occurrences of the first offset value satisfies a count threshold, accept the first offset value into a cluster; a weighted averager to calculate a weighted average of a second offset value and the first offset value in the cluster; a direction determiner to determine an origin direction of the media based on the weighted average; and a code reader to, when the origin direction is within a threshold angle, monitor audio codes embedded in the media.
1. An apparatus comprising:
an audio sample selector to select first audio samples of a first audio signal received by a first microphone, the first audio signal associated with media;
an offset selector to select first offset samples corresponding to a second audio signal received by a second microphone, the second audio signal associated with media, the first offset samples offset from the first audio samples by a first offset value;
a cluster analyzer to, if a first count of occurrences of the first offset value satisfies a count threshold, accept the first offset value into a cluster;
a weighted averager to calculate a weighted average of a second offset value and the first offset value in the cluster;
a direction determiner to determine an origin direction of the media based on the weighted average and
a code reader to, when the origin direction is within a threshold angle, monitor audio codes embedded in the media.
2. The apparatus of claim 1 , wherein the first count is a number of times the first offset value is selected by the offset selector.
3. The apparatus of claim 1 , wherein the offset selector selects the first offset samples based on a correlation value between the first audio samples and the first offset samples.
4. The apparatus of claim 1 , wherein:
the offset selector is to select second offset samples based on a correlation value between second audio samples and the second offset samples, the second offset samples corresponding to the second audio signal, the second offset samples offset from the second audio samples by the second offset value; and
the cluster analyzer is to accept the second offset value if a second count of occurrences of the second offset value satisfies the count threshold.
5. The apparatus of claim 4 , wherein the weighted averager is further to weight the first offset value based on the first count and weight the second offset value based on the second count.
6. The apparatus of claim 1 , wherein the count threshold is based on a difference between the first offset value and a third offset value in the cluster.
7. The apparatus of claim 1 , further including a signature generator to generate a signature when the origin direction of the media is within the threshold angle, the signature being representative of a characteristic of the media.
8. A non-transitory computer readable medium comprising instructions which, when executed, cause a processor to at least:
select first audio samples of a first audio signal received by a first microphone, the first audio signal associated with media;
select first offset samples corresponding to a second audio signal received by a second microphone, the second audio signal associated with media, the first offset samples offset from the first audio samples by a first offset value;
if a first count of occurrences of the first offset value satisfies a count threshold, accept the first offset value into a cluster;
calculate a weighted average of a second offset value and the first offset value in the cluster;
determine an origin direction of the media based on the weighted average; and
when the origin direction is within a threshold angle, monitor audio codes embedded in the media.
9. The non-transitory computer readable medium of claim 8 , wherein the first count is a number of times the first offset value selected.
10. The non-transitory computer readable medium of claim 8 wherein the first offset samples are selected based a correlation value between the first audio samples and the first offset samples.
11. The non-transitory computer readable medium of claim 8 , wherein the instructions, when executed, further cause the processor to:
select second offset samples based on a correlation value between second audio samples and the second offset samples corresponding to the second audio signal, the second offset samples offset from the second audio samples by the second offset value; and
accept the second offset value if a second count of occurrences of the second offset value satisfies the count threshold.
12. The non-transitory computer readable medium of claim 11 , wherein the instructions, when executed, further cause the processor to:
weight the first offset value based on the first count; and
weight the second offset value based on the second count.
13. The non-transitory computer readable medium of claim 8 , wherein the count threshold is based on a difference between the first offset value and the second offset value.
14. The non-transitory computer readable medium of claim 8 , wherein the instructions, when executed, further cause the processor to generate a signature when the origin direction of the media is within the threshold angle, the signature being representative of a characteristic of the media.
15. An apparatus to distinguish local media from spillover media comprising:
an audio sampler to sample a first audio signal from an audio source received by a first microphone and sample a second audio signal received by a second microphone from the audio source, the second microphone to be separated from the first microphone by a first distance;
a direction detector to calculate an origin direction of the audio source based on a time difference between receipt of the first and second sampled audio signals, the first distance, and a speed of sound; and
a code reader to, when the origin direction is within a threshold angle, detect an audio code embedded in at least one of the first sampled audio signal or the second sampled audio signal, the audio code to be representative of media corresponding to the first and second audio signals, the code reader to ignore the media when the origin direction exceeds the threshold angle.
16. The apparatus of claim 15 , wherein the audio source is a media device.
17. The apparatus of claim 15 , wherein the threshold angle is based on a physical location of the audio source.
18. The apparatus of claim 15 , wherein the threshold angle is based on an orientation of the apparatus and a physical location of the audio source.
19. The apparatus of claim 15 , further including a threshold processor to:
remove first samples from the first sampled audio signal based on intensities of the second samples being below a first threshold, the first threshold determined so that a first percentage of the sampled audio signal is below the first threshold; and
remove second samples from the second sampled audio signal based on intensities of the second samples being below a second threshold, the second threshold determined so that a second percentage of the sampled audio signal is below the second threshold.
20. The apparatus of claim 15 , further including a signature generator to generate a signature when the origin direction is within the threshold angle, the signature being representative of a characteristic of media corresponding to the first and second audio signal.