Google launches Gemini 2.0 Flash for photo editing

decrypt.co

Google has introduced a new version of its AI tool called Gemini 2.0 Flash. This tool allows users to edit photos simply by using words. It was launched last week and is now available to everyone after being tested for a year. Unlike many current photo editing tools, Gemini 2.0 Flash can modify existing images instead of just creating new ones. It understands both images and text at the same time, which helps it make specific changes while keeping the original content intact. Google explains that this capability leads to better illustrations, allowing users to tell stories with consistent characters and settings. Gemini’s design is different from competitors like OpenAI’s ChatGPT, which relies on multiple models for different tasks. Instead, Gemini 2.0 uses one model to handle both text and images. This could make it easier for users to get the results they want. When users tested the model, they found it could realistically modify photos. For instance, when one user asked it to add muscles to their self-portrait, it succeeded while keeping their face recognizable. However, the model has restrictions and does not allow edits involving children or nudity. Gemini 2.0 also performs well when changing art styles. Users could transform images into various styles, like manga or paintings. However, when asked to imitate specific artists, it sometimes instead reproduced existing artworks instead of adapting styles. For practical edits, the AI effectively removes or adds objects in images. Users found it could even replace a basketball with a rubber chicken. However, it sometimes altered unrelated parts, which could be corrected with other editing tools. The AI shows an impressive ability to shift perspectives, however, these changes are often new creations rather than direct edits. This allows the AI to understand three-dimensional space better, though it may not always get background changes right. Google is making this experimental tool available to developers and users who prefer not to share data with the company. Overall, Gemini 2.0 Flash is a notable new option for fun and creative image editing.


With a significance score of 2.9, this news ranks in the top 24% of today's 18142 analyzed articles.

Get summaries of news with significance over 5.5 (usually ~10 stories per week). Read by 9000 minimalists.


loading...