Video similarity detection method, apparatus, and device

1 . A video similarity detection method, comprising:

receiving, by a detection apparatus, a first video, and determining a key frame of the first video based on the first video;

determining, by the detection apparatus, a similar key frame and a second video, the similar key frame being determined based on the key frame, the similar key frame is located in the second video; and

inputting, by the detection apparatus, the key frame and the similar key frame to an editing type recognition model to obtain an editing type, the editing type recognition model is used to obtain the editing type for editing between the key frame and the similar key frame, conversion between the key frame and the similar key frame is implemented through editing by using the editing type, the editing type comprises one or more of: cropping, stitching, rotating, mirroring, blurring, text adding, icon adding, color changing, brightness changing, or contrast changing.

2 . The method according to claim 1 , the method further comprising:

outputting, by the detection apparatus, the second video or information about the second video to a display, the information about the second video comprising a name of the second video; and

outputting, by the detection apparatus, the editing type to the display.

3 . The method according to claim 1 , wherein the editing type recognition model comprises a first feature extraction branch, a second feature extraction branch, and a predictor, and the inputting the key frame and the similar key frame to the editing type recognition model to obtain the editing type comprises:

inputting, by the detection apparatus, the key frame to the first feature extraction branch, and inputting the similar key frame to the second feature extraction branch;

performing, by the first feature extraction branch, feature extraction on the key frame, and outputting an editing feature of the key frame;

performing, by the second feature extraction branch, feature extraction on the similar key frame, and outputting an editing feature of the similar key frame; and

inputting, by the detection apparatus, the editing feature of the key frame and the editing feature of the similar key frame to the predictor, and outputting, by the predictor, the editing type.

4 . The method according to claim 1 , the method further comprising:

inputting, by the detection apparatus, the key frame into the editing type recognition model;

receiving, by the detection apparatus, an editing feature of the key frame from the editing type recognition model;

inputting, by the detection apparatus, the similar key frame into the editing type recognition model;

receiving, by the detection apparatus, a similar key editing feature of the similar key frame from the editing type recognition model;

obtaining, by the detection apparatus, the editing type based on the editing feature of the key frame and the similar key editing feature of the similar key frame.

5 . The method according to claim 1 , wherein the method further comprises:

calculating, by the detection apparatus, a similarity between the first video and the second video; and

outputting, by the detection apparatus, the similarity to the display.

6 . The method according to claim 1 , wherein the determining the similar key frame and the second video based on the key frame comprises:

querying, by the detection apparatus, a video library based on the key frame to obtain the similar key frame from the video library; and

determining, by the detection apparatus, the second video based on the similar key frame.

7 . The method according to claim 1 , wherein the editing type recognition model is based on a neural network model.

8 . A detection apparatus, comprising:

a communications interface;

a non-transitory memory storing instructions; and

at least one processor in communication with the communications interface and the memory, the at least one processor configured, upon execution of the instructions, to perform the following steps:

receive a first video, and determine a key frame of the first video based on the first video;

determine a similar key frame and a second video, the similar key frame being determined based on the key frame, the similar key frame is located in the second video; and

obtain an editing type based on the key frame and the similar key frame, conversion between the key frame and the similar key frame is implemented through editing by using the editing type, the editing type comprises one or more of: cropping, stitching, rotating, mirroring, blurring, text adding, icon adding, color changing, brightness changing, or contrast changing.

9 . The apparatus according to claim 8 , the at least one processor further executing the instructions to:

output the second video or information about the second video to a display, the information about the second video comprising a name of the second video; and

output the editing type to the display.

10 . The apparatus according to claim 8 , the at least one processor further executing the instructions to:

receive the key frame;

perform feature extraction on the key frame;

output an editing feature of the key frame;

receive the similar key frame;

perform feature extraction on the similar key frame;

output an editing feature of the similar key frame; and

obtain the editing type based on the editing feature of the key frame and the editing feature of the similar key frame.

11 . The apparatus according to claim 8 , the at least one processor further executing the instructions to:

input the key frame to the first feature extraction branch and input the similar key frame to the second feature extraction branch;

perform feature extraction on the key frame and output an editing feature of the key frame;

perform feature extraction on the similar key frame and output a similar key frame editing feature of the similar key frame; and

input the editing feature of the key frame and the editing feature of the similar key frame to the predictor, and output, by the predictor, the editing type.

12 . The apparatus according to claim 8 , the at least one processor further executing the instructions to:

calculate a similarity between the first video and the second video; and

output the similarity to the display.

13 . The apparatus according to claim 8 , the at least one processor further executing the instructions to:

query a video library based on the key frame to obtain the similar key frame from the video library; and

determine the second video based on the similar key frame.

14 . The apparatus according to claim 8 , wherein the editing type recognition model is based on a neural network model.

15 . A non-transitory computer-readable media storing computer instructions for video similarity detection, that when executed by one or more processors, cause the one or more processors to perform the steps of:

receiving a first video and determining a key frame of the first video based on the first video;

determining a similar key frame and a second video, the similar key frame being determined based on the key frame, the similar key frame is located in the second video; and

inputting the key frame and the similar key frame to an editing type recognition model to obtain an editing type, the editing type recognition model is used to obtain the editing type for editing between the key frame and the similar key frame, conversion between the key frame and the similar key frame is implemented through editing by using the editing type, the editing type comprises one or more of: cropping, stitching, rotating, mirroring, blurring, text adding, icon adding, color changing, brightness changing, or contrast changing.