Video production is a professional skill, from removing unwanted backgrounds, generating video materials, to blurring faces to ensure privacy, every step symbolizes professionalism and time. But today, I want to introduce you to an AI tool that has completely changed my perception: “Runway”.
This article will take readers through the powerful features of Runway in a “hands-on” manner (even beginners can easily operate it). Let’s first take a look at the original video used for testing (showing some already generated magic tricks).
▋Green Screen Background Removal
Stop worrying about tedious background removal. Runway makes this task super easy; with just a few clicks, you can get a perfect background-free image.
You can see that there are some details that haven’t been handled well in the final product, but this can be continuously fine-tuned in the editing interface; here, the focus is on demonstrating the functionality.
▋Blur Faces
We often see interviewees’ faces pixelated in the news, and now Runway provides this feature to solve it with just one click!
Get the result
However, you will notice that when the hand is very close to the face, the pixelation disappears momentarily, which is part that needs fine-tuning.
▋Inpainting – Remove Objects from Video
Isn’t it annoying when unwanted objects appear in the video? In severe cases, the entire video may need to be reshot!
But with Runway, all of this can be solved with just a few clicks.
Then the object above really disappears!? If the image is relatively static, the effect will be better because, in this demo, you will notice some unnatural blurriness when the hand approaches the position of the object.
▋Video to Video – Re-render Video with Specific Materials
Do you want to express an existing video in a different form? Just provide the materials and it can generate for you!
This uses Gen-1, which generates video from video.
After providing the video, there are 3 ways on the right side to change the style, and you can use “Preview styles” to see if it meets your expectations.
➤ Image: Provide your own image.
➤ Presets: There are some preset templates for you to use.
➤ Prompt: Use text commands (using magicain)
Here, let me show you the effect generated using an image (does it have a Kamen Rider feel?).
▋Conclusion
After trying it out, I feel that the details of the images are not handled very well; however, if you are a beginner like me, just needing to use a few features inside (e.g., background removal, face blurring, object removal), I think the difficulty of getting started is close to zero.
WeChat ID｜HELLO Programmer
Video ID｜iron.guo