Adobe to Launch AI Video Creation Tool Later This Year
Adobe to launch Adobe Firefly Video Model, a generative AI-powered video creation tool. The tool can generate a five-second clip based on single prompts and interpret text and image inputs. Adobe clarifies that the tool is trained on public domain or licensed content from the Adobe Stock database.
The software giant aims to enhance its range of applications tailored to creative professionals with this innovative tool.
Named Adobe Firefly Video Model, this artificial intelligence tool will be launched in beta, expanding Adobe's Firefly suite, which currently includes image-generating applications for still images, designs, and vector graphics.
This move will position Adobe in the competitive market for AI-based video generation tools, where it will join the likes of OpenAI's Sora, Stability AI's Stable Video Diffusion, and other AI video applications developed by smaller startups.
According to Alexandru Costin, Adobe's vice president of generative AI, the tool can create a five-second clip based on a single prompt, interpreting both text and image inputs. Users will have the flexibility to specify camera angles, panning, motion, and zoom preferences.
Costin highlighted Adobe's commitment to meeting videographers' expectations regarding quality and prompt understanding. He emphasised the tool's ability to adhere closely to user prompts, surpassing other AI video models in responsiveness.
Adobe clarified that the video model is trained on public domain or licensed content, ensuring compliance with intellectual property regulations. Costin specified that the training data is sourced exclusively from the Adobe Stock database, which comprises 400 million images, illustrations, and videos devoid of intellectual property, trademarks, or recognisable characters.
In addition to the Adobe Firefly Video Model, Adobe is set to introduce Generative Extend, a feature within its Premiere Pro video editing software. This tool can automatically extend any existing clip by two seconds by generating suitable inserts to seamlessly fill gaps in the footage.
Costin expressed enthusiasm about the positive reception of Generative Extend since its preview in April, noting the widespread approval from Adobe's customer base.
Adobe to launch Adobe Firefly Video Model, a generative AI-powered video creation tool
The tool can generate a five-second clip based on single prompts and interpret text and image inputs
Adobe clarifies that the tool is trained on public domain or licensed content from the Adobe Stock database
Source: REUTERS