12 ways to remove duplicates in short videos! Increase originality by 80%! A must-see for newbies!

The current short video involution is really too serious, and the platform's review is becoming more and more strict . The most common problem encountered by people who make videos is "the repetition is too high, and the recommendation will be affected".

It seems that only by being original can there be a way out. Let’s not talk about whether everyone has the energy or not. Why can others be able to successfully carry it?

In fact, there are still technical factors. I can't do original work and I don't want to appear in real people. I just want to simply download the material on the Internet and edit it a little to make it my own work. This step must be taken seriously.

The more thorough the redo, the higher your originality, and the video will naturally not be detected as too repetitive.

First of all, we need to know that the short video is transported and deduplicated. It is not enough to change the MD5 and add a filter. Now the platform repeatedly detects the short video from multiple dimensions. Simply changing the MD5 is useless, because you do anything to the video. operation, MD5 will change accordingly, there is no need to mention it separately.

The filter is nothing more than modifying the color value. The effect is not much different from that of MD5. For example, if you add a filter to the frame detection, the effect is the same.

So if you want to be a qualified porter, how do you need to process the video so that it will not be detected by major platforms? The next brother will focus on sharing, skills and processing methods for short videos to those who need it.

Platform check mechanism

Let’s briefly talk to everyone about the platform’s duplicate checking mechanism. After understanding how the platform detects duplicates, we can know how to deal with them and how to remove duplicates.

1. The platform will first check MD5, because this check is the fastest

2. Then check the title, label, cover, description, duration, resolution, bgm and other information of the video, and focus on some possible duplicate videos screened out at this stage. Duplicate detection like these attributes is relatively simple .

We need to know that as far as the current technology is concerned, there is no way for the video to directly compare whether it is the same or not, unless it is checked manually.

However, comparing the similarity of pictures is relatively simple and fast, so comparing the similarity of two videos is actually a comparison of pictures.

Generally, the platform will extract a few frames from the beginning and end of the video, and then extract a few frames in the middle for comparison. If the picture changes are relatively small, the comparison of the basic pictures will be the same. (This is just a general principle, of course, the actual detection is definitely not so simple)

We know the principle of detection. It is relatively simple to achieve de-duplication. Next, Zhunge will talk about how to de-duplicate in detail. Let me give a few examples. In fact, different types of videos have different de-duplication methods.

12 Deduplication Methods

1. Watermark: remove the watermark of the original video, and then add the exclusive watermark of your own account ;

2. Trim: Trim the size and ratio of the original video to form a new video ratio ;

3. Mirror: Use the mirror function to change the screen ;

4. Titles and endings: It can't be the same as the original video, first cut the beginning and ending, and then use editing software to re-create the title and ending, it doesn't need to be too long, 3-5 seconds is enough ;

5. Special effects: add more special events and special effects ;

6. Subtitles: Change the picture effect by enlarging the subtitles through the background and the picture;

7. Variable speed: use the variable speed function to slow down or fast forward to change the overall speed of the material;

8. Change the background: use the intelligent keying function to directly change the background of the screen;

9. Frame rate: Some platforms need to specify 60fps. If it is lower than this number, it cannot be published, then the frame rate needs to be modified. There are also many software for modifying the frame rate and resolution, such as "Format Factory".

10. Resolution: For example, the original video is 1920x1080 Blu-ray resolution, then you can change it to 720x1280;

11.BMG: The volume also needs to be adjusted, you can turn it up or down, it is best to be able to dub and comment by yourself;

12. Ultra-clear image quality: Select high-quality export, and use the ultra-clear image quality function of Quick Shadow Treasure Box to process the video to 4K.

E.g

The more common way to change the screen is to switch between horizontal and vertical screens by inserting background pictures, background colors, screen overlays or blurred borders.

Look at the picture below, the first picture inserts a black background, the second inserts a blurred edge, the third inserts a picture, and the fourth inserts a picture overlay.

Videos such as movie commentary belong to the second cut, because their own subtitles and explanations are added. This kind of video is an original video by the platform by default.

The thing that is completely different from the mixed editing of film and television is that the film and television are copyrighted. Even if the mixed editing of film and television adds a bgm .

If you want to use someone else's commentary video, re-dubbing and adding subtitles yourself, you need to pay attention to the following points. The length of the original video cannot be the same. The purpose of emphasizing the different length is that when the same moment is played, the picture is different. The frame will not be detected, and remember to add your own intro and credits.

The simplest original is to synthesize a video from a picture, and then add a bgm. It is completely fine to say that it is original, unless someone has synthesized the same video before you.

Summary: Two Dimensions of Deduplication

1. Video external information, including title, cover, description and tags ;

2. Information about the video itself, including material content, resolution, video duration, bgm, etc.

Core: Change the video screen and underlying parameters as much as possible without changing the look and feel of the original video too much.

Although the review mechanism of each platform is getting stricter and stricter, you don't have to worry too much. If you clarify the operating rules and review mechanism of the platform, you will definitely have some fun.

The deduplication methods shared above can all be operated in batches. Needless to say, vlogs cannot be done in batches.

However, it is better to keep pace with the times in handling. The platform algorithm is constantly being upgraded, and the deduplication method also needs to be constantly improved. The easiest way is to learn how others do it.