Alibaba Introduces Open-Source AI Model To Create And Play Videos

JAKARTA Alibaba launched a new open-source model called Wan2.1-VACE (Video All-in-on Creation and Editing). As the name implies, this model was released to create and edit videos.

VECE is the industry's first open-source model specifically designed to create and edit videos. This Artificial Intelligence (AI) model supports the creation of videos with various input models, ranging from text, images, to videos.

According to Alibaba's claim, VACE provides advanced editing features. Creators on various social media can take advantage of this technology to look for image and frame references, do repainting videos, to expand Spatio-temporal.

With high level flexibility, this model can even turn on static images by adding a natural movement effect. In addition, there is an option to change poses, adjust movement, image depth, to easily change the video color.

Therefore, VACE is referred to as an all-in-one model. In fact, this open-source model is capable of adding, modifying, and removing certain areas in the video. All these capabilities will not affect the surrounding part.

Wan2.1-VACE also offers other functionalities such as turning a stationary image into a video with the direction of the adjustable object. Users can change the character or object in the video smoothly.

In addition, users can minimize the characters they refer to, control the pose easily, and expand the vertical image to horizontal. This model can be downloaded for free on the Hugging Facte, GitHub, and the ModelScope community.