Sora's New Capabilities: Video Inpainting and Integration into Adobe's Premiere

🎥Sora can inpaint videos to a certain degree, allowing users to swap out prompts and change focal characters.

🖼️Sora has the potential for image to video generation, as seen in an example of an AI-generated dog.

✂️Sora can extend video clips in either direction, as demonstrated with a train car video.

🎬Adobe's recent integration of Firefly video and Pika Runway suggests the possibility of Sora integration into Premiere.

💡Paul Trilo's music video showcases Sora's surreal and dreamy visuals, utilizing straight cuts instead of VFX.

Can Sora generate image to video?

—Yes, Sora has the potential for image to video generation, as shown in an example of an AI-generated dog wearing a beret.

What other capabilities does Sora have?

—Sora can inpaint videos, allowing users to change prompts and focal characters. It can also extend video clips and potentially integrate into Adobe Premiere.

Does Sora have character consistency?

—No, character consistency is still a challenge for Sora, as seen in the varying appearances of characters throughout a video.

How long does it take for Sora to generate videos?

—The generation time for Sora can range from 15 minutes to 1 hour, depending on the amount and length of clips used.

What is the visual style of Sora?

—Sora's visuals have a gritty and surrealistic quality, as seen in Paul Trilo's music video, which embraces the weirdness.

02:15Sora showcases its video inpainting capabilities by swapping out prompts and changing focal characters.

02:35Sora has the potential for image to video generation, as demonstrated with an AI-generated dog wearing a beret.

03:03Sora can extend video clips in either direction, demonstrated through a train car video.

04:13Adobe's integration of Firefly video and Pika Runway suggests the possibility of Sora integration into Premiere.

05:29Paul Trilo's music video showcases Sora's surreal and dreamy visuals, without relying on additional VFX.