Earlier today we launched an advanced image generation and editing model, available in the Gemini app, AI Studio and Vertex AI. This update introduces significant progress in character consistency; Exactly, conversation editing; And the ability to combine photos into a whole new creation. To help you get the most out of this update, here are some tips to write more effective requests for image generation and editing in Gemini.
Key features in image generation in Gemini
Before you dive in, it is useful to make you familiar with what has been improved in Gemini, so you can consider which use cases to try with it:
- Consistent character design. Preserve a character or object appearance across multiple generations and edits.
- Creative composition. Mix different elements, topics and styles from multiple concepts to a single, overall image.
- Local edits. Make precise edits for specific parts of an image using simple language.
- Design and appearance adjustment. Apply one style, texture or design from one concept to another.
- Logic and reasoning. Use understanding in the real world to generate complex scenes or predict the next step in a sequence.
6 items in the construction of effective prompts
You can get good results with Gemini from simple one or two-phrasing inputs. To achieve the best results and unlock more nuanced creative control, consider including the following items in your prompt:
- Subject: Who or what is in the picture? Be specific. (eg, A stoic robotic barista with glowing blue optics; A fluffy calico -cat wearing a little wizard).
- Composition: How is the shot framed? (eg, Extremely close upAt Wide shotAt Low angular imageAt Portrait).
- Action: What’s going on? (eg, Brewing a cup of coffeeAt To throw a magic spell formAt in the middle of the strike runs through a field).
- Location: Where does the scene take place? (eg, A futuristic café on MarsAt A messy Alchemists LibraryAt A sunshine meadow at Golden Hour).
- Style: What is the total aesthetics? (eg, 3D animationAt Film NoirAt Watercolor paintingAt PhotorealisticAt The 1990s product photography).
- Editing Instructions: To change an existing image, you need to be direct and specific. (eg, Change the man’s tie to greenAt Remove the car in the background).
Springing Examples: A showcase with creative techniques
Different encouragement strategies can lock everything from photo -realistic edits to amazing new worlds. Here are five techniques to try, each with an important example.
1. Preserve the characters’ appearances.
Gemini can maintain resemblance to a person or character across different positions, lighting and environments and even apply the same character to new styles and surfaces. Here is an example of how a character can be used across multiple prompts in the same session:
- Quick 1: A quirky illustration of a small, glowing mushroom sprite. Sprite has a large, bioluminescent fungal jacket for a hat, wide, curious eyes and a body made of woven vines.
- Prompt 2 (in the same conversation): Now show the same sprite running on the back of a friendly, moss -covered snail through a sunny meadow full of colorful wild flowers.