Methods and apparatus to detect dynamic ad insertion misalignment

1 . An apparatus to identify advertisement misalignment, the apparatus comprising:

interface circuitry;

machine-readable instructions; and

programmable circuitry to at least one of instantiate or execute the machine-readable instructions to:

identify a timestamp using a manifest file of a video stream, the timestamp corresponding to a start of a dynamic advertisement insertion (DAI);

form a list of consecutive video segments, the list including at least a content segment that ends at the timestamp and a DAI segment that begins at the timestamp;

determine that a transition from the content segment to the DAI segment fails to satisfy a quality threshold, the failure based on a determination the transition does not include aligned black frames, a determination the transition does not include a low priority bumper sequence, and a determination the transition is misaligned with a scene change; and

flag the timestamp as a misalignment.

2 . The apparatus of claim 1 , wherein the instructions cause the programmable circuitry to notify a content provider in response to the flag, the content provider to perform operations to correct the misalignment.

3 . The apparatus of claim 1 , wherein to determine the transition does not include aligned black frames, the instructions cause the programmable circuitry to:

determine a first frame after the timestamp is not black;

determine a last frame before the timestamp is not black; and

determine a the list of video segments does not include a black frame within a threshold time range of the timestamp.

4 . The apparatus of claim 1 , wherein to determine the transition does not include a low priority bumper sequence, the instructions cause the programmable circuitry to:

count a number of consecutive frames that are similar images to one another; and

determine the number is less than a threshold value.

5 . The apparatus of claim 1 , wherein to determine the transition does not include a low priority bumper sequence, the instructions cause the programmable circuitry to:

identify a bumper sequence by counting a number of consecutive frames that are similar images to one another;

determine the bumper sequence includes a logo; and

determine a size of the logo relative to a size of the corresponding frame is above a threshold.

6 . The apparatus of claim 5 , wherein the instructions cause the programmable circuitry to execute a machine learning model to perform one or more of:

determine whether consecutive frames are similar images to one another;

determine the bumper sequence includes a logo; and

determine a size of the logo relative to a size of the corresponding frame is above a threshold.

7 . The apparatus of claim 1 , wherein:

the timestamp is a first timestamp;

to determine the scene change is misaligned with the transition, the instructions cause the programmable circuitry to:

identify, based on image data within the list of video segments, a scene change at a second timestamp; and

determine a difference between the first timestamp and the second timestamp is greater than a threshold.

8 . The apparatus of claim 1 , wherein the

the timestamp is a first timestamp;

to determine the scene change is misaligned with the transition, the instructions cause the programmable circuitry to:

identify, based on image data within the list of video segments, a scene change at a second timestamp;

determine a difference between the first timestamp and the second timestamp is less than a threshold; and

determine that audio levels at the second timestamp exceed a threshold.

9 . The apparatus of claim 1 , wherein:

the list of video segments is a first list of video segments;

the transition is a first transition; and

the instructions cause the programmable circuitry to:

form a second list of consecutive video segments corresponding to a second timestamp where the DAI ends; and

determine whether a second transition from an end of the DAI to a resumption of the content segment satisfies a quality threshold.

10 . The apparatus of claim 1 , wherein the instructions cause the programmable circuitry to provide statistics regarding the transition from the content segment to the DAI segment in response to flagging the transition as a DAI misalignment.

11 . A non-transitory machine-readable storage medium comprising instructions to cause programmable circuitry to at least:

identify a timestamp using a manifest file of a video stream, the timestamp corresponding to a start of a dynamic advertisement insertion (DAI);

form a list of consecutive video segments, the list including at least a content segment that ends at the timestamp and a DAI segment that begins at the timestamp;

determine that a transition from the content segment to the DAI segment fails to satisfy a quality threshold, the failure based on a determination the transition does not include aligned black frames, a determination the transition does not include a low priority bumper sequence, and a determination the transition is misaligned with a scene change; and

flag the timestamp as a misalignment.

12 . The non-transitory machine-readable storage medium of claim 11 , wherein the instructions cause the programmable circuitry to notify a content provider in response to the flag, the content provider to perform operations to correct the misalignment.

13 . The non-transitory machine-readable storage medium of claim 11 , wherein to determine the transition does not include aligned black frames, the instructions cause the programmable circuitry to:

determine a first frame after the timestamp is not black;

determine a last frame before the timestamp is not black; and

determine a the list of video segments does not include a black frame within a threshold time range of the timestamp.

14 . The non-transitory machine-readable storage medium of claim 11 , wherein to determine the transition does not include a low priority bumper sequence, the instructions cause the programmable circuitry to:

count a number of consecutive frames that are similar images to one another; and

determine the number is less than a threshold value.

15 . The non-transitory machine-readable storage medium of claim 11 , wherein to determine the transition does not include a low priority bumper sequence, the instructions cause the programmable circuitry to:

identify a bumper sequence by counting a number of consecutive frames that are similar images to one another;

determine the bumper sequence includes a logo; and

determine a size of the logo relative to a size of the corresponding frame is above a threshold.

16 . The non-transitory machine-readable storage medium of claim 11 , wherein:

the list of video segments is a first list of video segments;

the transition is a first transition; and

the instructions cause the programmable circuitry to:

form a second list of consecutive video segments corresponding to a second timestamp where the DAI ends; and

determine whether a second transition from an end of the DAI to a resumption of the content segment satisfies a quality threshold.

17 . The non-transitory machine-readable storage medium of claim 15 , wherein the instructions cause the programmable circuitry to execute a machine learning model to perform one or more of:

determine whether consecutive frames are similar images to one another;

determine the bumper sequence includes a logo; and

determine a size of the logo relative to a size of the corresponding frame is above a threshold.

18 . The non-transitory machine-readable storage medium of claim 11 , wherein:

the timestamp is a first timestamp;

to determine the scene change is misaligned with the transition, the instructions cause the programmable circuitry to:

identify, based on image data within the list of video segments, a scene change at a second timestamp; and

determine a difference between the first timestamp and the second timestamp is greater than a threshold.

19 . A method to detect DAI misalignment, the method comprising:

identifying a timestamp using a manifest file of a video stream, the timestamp corresponding to a start of a dynamic advertisement insertion (DAI);

forming a list of consecutive video segments, the list including at least a content segment that ends at the timestamp and a DAI segment that begins at the timestamp;

determining, by performing operations with programmable circuitry, that a transition from the content segment to the DAI segment fails to satisfy a quality threshold, the operations with programmable circuitry to include:

determining the transition does not include aligned black frames;

determining the transition does not include a low priority bumper sequence; and

determining the transition is misaligned with a scene change; and

flagging the timestamp as a misalignment.

20 . The method of claim 19 , wherein determining the transition does not include aligned black frames includes:

determining a first frame after the timestamp is not black;

determining a last frame before the timestamp is not black; and

determining a the list of video segments does not include a black frame within a threshold time range of the timestamp.