[ad_1]
Adobe last year launched Firefly, its latest generative AI model based on its predecessor SenseiAI, and now the company is showing how its video editing app Premiere Pro will be used. In an early look, it showed off some of the key features that will arrive later this year, including adding and removing objects, generative expansion, and converting text to video.
The new features are likely to be very popular, since cleaning up video is a common (and painful) task. The first feature, Geneative Extend, addresses the problem editors face with almost every edit: clips that are too short. “Add frames seamlessly to make clips longer, so it’s easier to time your editing perfectly and add smooth transitions,” says Adobe. This is done by using artificial intelligence to create additional media, which helps with editorial or transmission coverage.
Another common problem is unwanted objects that you don’t want in the shot which can be difficult to remove or add your own Do Wants. Adding and removing objects in Premiere Pro addresses this, again using Firefly’s generative AI. “Simply select objects, track them, and then replace them. Remove unwanted items, change an actor’s wardrobe, or quickly add dressings like a painting or realistic flowers on a desk,” Adobe wrote.
Adobe shows some examples, where you add a pile of diamonds to your briefcase via a text prompt (generated by Firefly). It also removes an ugly toolbox, changes the clock face, and adds a bow tie to the character’s costume.
The company also demonstrated a way in which it can import custom AI models. One, called Pika, is what powers Geneative Extend, while the other (OpenAI’s Sora) can automatically generate B-Rolls. Adobe said in the video that the latter is bound to be controversial because it could potentially eliminate thousands of jobs, but it is still “currently in the early research stage.” The company notes that it will add “content credentials” to such snapshots, so you can see what was created by AI including the company behind the model.
A similar feature is also available in Text to Video, allowing you to create brand new footage directly within the app. “Simply type text into a prompt or upload reference images. These clips can be used to visualize and create storyboards, or to create B-roll to augment live-action footage,” Adobe said. The company appears to be commercializing this feature very quickly, considering that the generative AI video debuted just a few months ago.
These features will arrive later this year, but Adobe will also offer updates to all users in May. This includes interactive fade handles for easier transitions, and a basic audio badge with audio category tagging (the AI automatically tags audio clips as dialogue, music, sound effects, or ambience, and adds a new icon so editors have instant one-click access to the controls “correct for the task”), influence badges, and redesigned waveforms in the timeline.
[ad_2]
Source link
