Adobe's New Tool: Is Adobe About to Change Video Creation

As part of its effort to strengthen its portfolio of products for creative professionals, software giant Adobe announced on Wednesday that it will launch a new generative AI-powered video creation and editing tool in a restricted release later this year. The artificial intelligence tool, Adobe Firefly Video Model, will be made available in beta form and added to the Photoshop creator's portfolio of Firefly image-generating programs, letting users create vector drawings, still photos, and designs.

Adobe's New Tool: Future Video Capabilities

Adobe's future video capabilities, which were first shown earlier this year, are intended to give film editors and video professionals the tools they need to inspire their creative vision, fill in timeline gaps, and add new features to existing material. With the addition of the Firefly Video Model, Adobe's family of generative AI models—which now consists of the Image, Vector, and Design models—will grow to become the most complete model suite for creative teams. Over 12 billion photographs have been created using Adobe Firefly worldwide to date.

Adobe's New Tool: New Firefly-powered Video and Image-to-Video

New Firefly-powered Video and Image to Video features will be accessible on Firefly.Adobe.com later this year. Premiere Pro, the top video editor in the business, will come pre-integrated with Generative Extend. Text-to-video features allow you to create videos based on text prompts, edit videos with a wide range of camera parameters (zoom, motion, and angle), and create a B-Roll that smoothly fills in the spaces in your timeline by referencing photos. The ability to convert still images or graphics into breathtaking live-action clips is an image-to-video skill.

The Firefly Video Model will support a wide range of use cases.

The Firefly Video Model will support a wide range of use cases, which is intended to assist video professionals in obtaining the ideal clip. Potential applications might be, for example, using fine-grained controls like camera motion and angle to create captivating animations and effects, using camera motion to create an insert into a crucial video shot, or combining video with expertly shot footage to tell stories more quickly. Editors can create with confidence knowing that the Firefly Video Model is made to be commercially safe and is only trained on the content we have permission to use—never on Adobe customer content. This is true of all Adobe Firefly generative AI models, including the Adobe Firefly Image 3 Model, Adobe Firefly Vector Model, and Adobe Firefly Design Model.

With this paradigm, Adobe will take a more prominent place in the expanding market for AI-based video creation tools, which is already being pursued by smaller businesses such as Sora from OpenAI, Stable Video Diffusion from Stability AI, and others.

Adobe's New Tool: Generative AI at Adobe

According to Alexandru Costin, vice president of generative AI at Adobe, the tool can understand both word and visual cues. It can create a five-second clip for a single query. Additionally, users can designate the necessary zoom, motion, panning, and camera angle.

"We've made the necessary investments to ensure that this model meets the high standards for clarity and speed that videographers demand. In an interview with Reuters, Costin stated, "We've invested in making sure we pay attention to the prompt... respecting guidance from videographers much better than other (AI video) models." According to Adobe, the video model was trained on licensed or public domain content granted permission to use rather than on content owned by Adobe customers.

Adobe is introducing Generative Extend.

"We only train them on the Adobe Stock database of content that contains 400 million images, illustrations, and videos that are curated not to contain intellectual property, trademarks or recognizable characters," Costin added. Additionally, Adobe is introducing Generative Extend, a feature that will be included in its Premiere Pro video editing program. It will allow users to add two seconds to any existing clip by creating a suitable insert to fill in any missing frames. According to Costin, the tool has received "a huge positive reaction from all of our customers" since its April preview.