Open-source AI video tool for all

Alibaba has unveiled Wan2.1-Vace, an open source Amnesty International model designed to shake how we create and edit videos.

VACE does not appear from delicate air. It is part of the Wan2.1 family wider than the Video AI family. They make a somewhat bold claim for that, saying that it is “the first open source model in the industry to provide a unified solution for various video and editing tasks.”

If Alibaba is able to succeed in converting users away from having to reconcile many tools, it will separate from one axis-it can be a real change in the game.

So, what can this thing really do? Well, for beginners, it can provoke videos using all kinds of claims, including text orders, fixed images, and even other videos.

But it is not only about making videos from scratch. The editorial set supports the reference to the specific photos or tires to direct artificial intelligence, advanced “re -paint” (more about it in a second), and switch only specific parts of the current video, and even the extension of the video. Alibaba believes these features “Enable a flexible mix of different tasks to enhance creativity.”

Imagine you want to create a video clip with specific characters interacting, perhaps based on some of the photos you have. Weiss claims that he is able to do so. Have you got a fixed picture that you wish was dynamic? An artificial intelligence model in alibaba can add a natural movement of appearance to return it to life.

For those who like to control performance, there are advanced “video repainting” functions that I mentioned earlier. This includes things such as transferring shapes from one topic to another, accurate control of movement, controlling the perception of depth, and even changing colors.

One of the features that caught my attention is its ability to “support addition, modification or deletion to selective areas from a video without affecting the ocean.” This is a huge feature for detailed adjustments – no more absolutely tampering with background when you only try to modify one small element. In addition, it can make your video fabric larger and fill the new space with relevant content to make everything look richer and more expanding.

You can take a flat picture, turn it into a video, and tell the objects exactly how to move by drawing a path. Do you need to switch a letter or an object with something else you offer as a reference? No problem. Moving those figures referred to? finished. Control their status specifically? I got it.

Alibaba gives an example of an open source AI model taking a long, slim vertical image and intelligently expanding it to an elabage video, adding new pieces and pieces automatically via the reference images or other claims. This is very elegant.

Of course, Fassi is not just a magic. There is some smart technology concerned, designed to deal with the dominant reality of video editing. The main piece is something called Alibaba Video Case Unit (VCU), which “supports the unified treatment of multimedia inputs such as text, photos, videos and masks.”

Then there is what they call the “context transformer structure”. This smart thing of engineering “injecting different tasks concepts using formal and spatial dimensions.” Basically, think about it gives artificial intelligence a good understanding of time and space within the video.

With all this smart technology, Alibaba believes that VACE will be a great success in a few regions. Think about fast social media clips, attractive ads, marketing content, special effects after the arduous production of the film and television, and even to generate educational videos and dedicated training.

Alibaba makes Wan2.1-Vace open source to spread the love of artificial intelligence

Building these artificial intelligence models usually costs a wealth and needs a huge computing power and tons of data. So, open source wan2.1-Vace? This is a big problem.

“Open access helps to reduce the barrier in front of more companies to take advantage of artificial intelligence, allowing them to create high -quality visual content specifically designed to meet their needs, quickly and effectively costly,” explains Ali Baba.

Basically, Alibaba hopes to allow more people-especially smaller and creative companies-by placing their hands on the first-class AI without breaking the bank. This democracy for strong tools is always a welcome scene.

They do not just fall one version. There is a huge parameter model of 14 billion people for those who have a dangerous horse power, and a more intelligent teacher with a capacity of 1.3 billion lighter settings. You can now seize it for free on Hugging Face and GitHub, or via the Alibaba Cloud community of open sources, Modelscope.

(Photo source: www.alibabagroup.com)

See also: The United States closes the brakes on the basis of the spread of artificial intelligence

Do you want to learn more about artificial intelligence and large data from industry leaders? Check AI and Big Data Expo, which is held in Amsterdam, California, and London. The comprehensive event was identified with other leading events including the smart automation conference, Blockx, the digital transformation week, and the Cyber ​​Security & Cloud.

Explore the upcoming web events and seminars with which Techforge works here.