Google launches AI image creation tool from real images

(CLO) Google has just launched a new artificial intelligence (AI) tool called "Whisk", allowing users to upload photos to retrieve AI-generated combined images, even if the user does not enter any text.

Whisk uses AI to combine the subject, context, and style of an uploaded photo, creating a new and more unique image.

Whisk is described by Google as a “creative tool” that helps users quickly generate new visual ideas without requiring professional photo editing skills. According to Google, the tool is not a traditional image editor, but rather a fun AI tool that aims to spark creativity and rapid discovery.

When a user uploads a photo, Whisk uses a combination of Google's AI service, Gemini, and Imagen 3 technology — a tool that creates images from text that Google acquired from DeepMind.

Gemini will analyze the photo and create a caption, then Imagen 3 will combine elements of that photo in a creative way, retaining the “essence” of the subject rather than copying it exactly.

Whisk tool interface. Screenshot

This means that the final result may not be 100% identical to the original image. For example, the height, hairstyle, or skin tone of subjects in the new image may be different from the original. However, users can still adjust the input information, change the scene, style, or combine multiple subjects to create different images.

Whisk can create images not only from text but also from raw images, expanding creative possibilities without requiring users to have any photo editing experience. "Whisk is designed to help users remix subjects, contexts, and styles in creative ways, allowing them to explore visually rather than meticulously edit each pixel," said Thomas Iljic, director of product management at Google Labs.

While Whisk is in its early stages of development, the tool has been launched as a website on Google Labs and is now available to users in the US.

Dan Ives, managing director and senior analyst at Wedbush Securities, said Whisk marks another “show of force moment” for Google in the tech race.

Ives also noted that DeepMind, the AI ​​lab Google acquired in 2014, is a key asset in helping Google maintain its position in the AI ​​space. AI products, including Whisk, are a key part of Google’s product development strategy in the coming years, with several new products expected to launch by 2025.

The Whisk tool opens up new avenues for using AI to create creative products with minimal user intervention. This shows the advancement of AI in understanding and combining visual elements creatively.

Whisk is part of a growing trend among major tech companies, including Google and OpenAI, to develop consumer-facing AI tools that aim to enable new creative experiences, from generating images, text, and videos. OpenAI recently introduced a text-to-video generator called Sora, which competes directly with Whisk.

Ngoc Anh (according to CNN, The Verge, ZDNET)

Source: https://www.congluan.vn/google-ra-mat-cong-cu-tao-hinh-anh-ai-tu-hinh-anh-that-post326441.html