Adobe launched on Thursday the latest version of his Firefly family of image generation models, a model to generate vectors and a redesigned web application That houses all its AI models, in addition to some of its competitors. There is also a mobile application for Firefly in process.
Adobe says that the new Firefly Image Model 4 improves its predecessors in terms of quality, speed and the amount of control over the structure and style of the outputs, the angles of the camera and the zoom. You can generate images in resolutions of up to 2k. There is also an adjusted and more capable version of this model called Image Model 4 that can make complex scenes that consist of small structures and many details.
Alexandru Costin, vice president of generative in adobe, said the company trained models with a superior order of calculating to allow them to generate more detailed images. He added that, compared to the previous generation models, the new models improve the generation of text in images and have characteristics that allow users to use images of their choice so that the model generates images in that style.
The company is also doing its Firefly video model, which Released in Beta Limited last yearAvailable for everyone. It allows users to generate video clips from a text or text image, use camera angles, specify start and completion frames to control the shots, generate atmospheric elements and customize movement design elements. The model can generate video clips from text in resolutions of up to 1080p.
Meanwhile, the Firefly vector model can create editable vector -based works, iterate and generate variations of logos, products of products, icons, scenes and patterns.
In particular, the Firefly web application gives access to all these models, as well as some models of image generation and videos of OpenAi (GPT Image Generation), Google (Image 3 and I see 2) and FLX (Flux 1.1 Pro). Users can change between any of these models anywhere, and the images generated from any model will have attached content credentials. The company hinted that other AI models can be added to the web application in the future.
Adobe is also publicly testing a new product called Firefly Boards, a canvas for ideation or mood. It allows users to generate or import images, remix them and collaborate with others, similar to what they can do with other idea based on Visual Electric, Cove either Space. Boards is available through Firefly’s web application.
The company said that these new models would soon be integrated into their product portfolio, but did not give a timeline for deployment.
Adobe is also making his text API in the image and Avatar’s API are generally available, and said that a new video to video API is now available for use in beta. These APIs are available through the API collection, Firefly Services of the company.
Adobe is also testing a web application called Adobe content authenticity Allow users to credentials to their work to indicate property and attribution through metadata. Users can also indicate whether IA companies can use images for the training of the AI model.
