Competition heats up as Microsoft and Adobe release AI image generation tools

With the launch of OpenAI DALL-E and Midjourney, AI art generators are starting to become more popular, taking text prompts and turning them into beautiful, often surreal artworks - and today, there are Two major companies have joined the ranks.

Microsoft announced that it will introduce the AI ​​image generation function powered by the DALL-E model to the Bing search engine and Edge browser through Bing Image Creator. Creative software developer Adobe also revealed that it will enhance its tools with an AI art generation product called Firefly.

This new AI image generator is already available in Creative mode for users who have access to Bing Chat Preview. Users get an interface that generates artwork for them by simply entering a description of the image they want to create, providing a rich example such as a place, thing, or activity, and the underlying AI creates based on that information.

Users can quickly and easily input their imagination, including "create image" or "draw image", and the AI ​​will take it from there, including by changing internal parts, changing image elements, allowing the user to modify the background or Other parts to iterate over previous images. Edge users with access rights can also use this feature by clicking the Bing Image Creator icon in the sidebar or calling the Bing chat function in the browser.

Microsoft emphasizes that this AI image generator has built-in protection measures to prevent the creation of unsafe or harmful images. If the user attempts to use the prompt to generate such an image, the generator will block and warn user. This AI-created image also places a watermark icon in the lower left corner, so that if the image is used, it indicates that it was generated by Bing Image Creator - although the watermark may be cropped off.

Yusuf Mehdi, corporate vice president and consumer chief marketing officer at Microsoft, said: “With these updates and more coming, our goal is to deliver more immersive experiences in Bing and Edge. Making finding answers and exploring the web more fun, useful and fun."

This feature is now rolling out to users with access to the new Bing and Edge AI preview features, and is available in English . Users who don't have access to the new Bing and Edge features can sign up for the waitlist, or can test it out now.

Adobe launches Firefly generative AI tool

Adobe calls Firefly "a family of generative AI models for creative expression" and will add it to the app , allowing users to unleash the power of AI art generation, starts with two tools: one that allows users to generate users, and another that generates text effects.

The new AI art generation tools will be directly integrated into Adobe’s existing Creative Cloud creative cloud tool suite, including Creative Cloud, Document Cloud, Experience Cloud and Adobe Express, allowing users to take advantage of these features.

The first feature in the beta version is much like DALL-E or Midjourney, which allows users to enter a text prompt and generate a series of images based on a "text-to-image" text description. The second is similar to WordArt and is a "text effect", which allows the user to type text on the screen and is then prompted to adjust its appearance, such as "covered in snow" or "looks like made of cake". Styles are applied to text typed on the screen.

Adobe also plans many other upcoming features that will allow users to take content they have already created in Adobe products and use text to describe what they want to modify or add. This is the power behind generative AI, which can further extend the capabilities of Adobe’s already powerful AI tools based on Adobe Sensei.

For example, Photoshop and Illustrator graphics editor users will be able to take an image they are already working on and select a portion of their digital art, then have AI modify a certain part of the image using context-aware prompts. You can take a photo of a house on the beach, select the house, enter "shell house" and let the AI ​​draw a version of the shell house. Either choose ocean and ask the AI ​​to add ships, or choose sky and add alien battles.

It can also be used to create new images based on the same color scheme and style of existing content, easily generating similar images. The same technology can be applied to video editing, as AI is able to change moods and even the weather.

David Wadhwani, President of Digital Media at Adobe, said: “Generative AI is the next evolution in AI-driven creativity and productivity, transforming the conversation between creators and computers into something more natural, intuitive, and Powerful stuff. With Firefly, Adobe will bring AI-driven generation of 'creative ingredients' directly into customers' workflows."

To eliminate creators' concerns that Adobe may use their copyrighted content To address this concern, Adobe said the first model is based on Adobe Stock images, publicly licensed content and images with expired copyrights, to ensure that any images generated are safe for commercial use.

Adobe also said it is developing a model to compensate creators whose works are used to train generative AI models, such as the one used in Firefly, similar to the one it uses for Adobe Stock. In addition, Adobe has also pioneered a global standard that allows creators to attach a "do not train" metadata tag to their works to tell AI not to use their works.

Starting today, the first two Firefly tools are available in public beta.

The above is the detailed content of Competition heats up as Microsoft and Adobe release AI image generation tools. For more information, please follow other related articles on the PHP Chinese website!