Below are some prompts you can use to generate images using AI imaging tools like Stable Diffusion, StarryAI, Adobe Firefly, LeonardoAI, and NightCafe. The prompts were created using google bard and youbot. Google Bard was re branded as google gemini in early 2024.
One powerful tool for generating inspiration and creativity is the use of visual prompts. By incorporating images into your writing process, you can unlock new perspectives, ideas, and storylines that you may not have otherwise considered. In this blog post, we will explore the power of image prompts for stable diffusion, using Google bard and YouBot.
Basic Concepts or Guidelines for Image Prompt
Objects: You can describe a specific object, such as a “cat” or a “house.” You can also describe a more abstract object, such as a “feeling” or a “concept.”
Scenes: You can describe a specific scene, such as a “cityscape” or a “forest.” You can also describe a more abstract scene, such as a “dream” or a “memory.”
People: You can describe a specific person, such as a “friend” or a “celebrity.” You can also describe a more abstract person, such as a “perfect partner” or a “future self.”
Events: You can describe a specific event, such as a “wedding” or a “party.” You can also describe a more abstract event, such as a “miracle” or a “breakthrough.”
Colors: You can specify the colors you want to see in the image. For example, you could say “a blue sky” or “a red flower.”
Styles: You can specify the style of the image you want. For example, you could say “realistic” or “cartoonish.”
Moods: You can specify the mood of the image you want. For example, you could say “happy” or “sad.”
A flat image generated using Starryai. Prompt was created using google bard.
Generating Image Prompts with YouBot
I asked youbot (from you.com) and google bard to generate five random prompts using the above inputs. The results were as below:
1. An abstract scene with the mood of “serenity”. This could include soft colors, like pastel blues and pinks, and an image of a calm ocean or a peaceful meadow.
2. A realistic image of a red house with a garden in front. There could be flowers of different colors in the garden, and the image could be set in a neighborhood or countryside.
3. An abstract person with the mood of “hope”. The person could be represented as a silhouette or a blurry, dreamlike figure, and surrounded by bright, uplifting colors like yellow and orange.
4. A cartoonish scene of a forest with a group of animals having a party. The image could include a bear playing music on a guitar, rabbits dancing, and birds singing.
5. A realistic image of a cityscape at night with the mood of “loneliness”. The image could have shades of blue and grey, and depict a lone figure walking down a deserted street.
Forest animals having a party. Image generated using Adobe Firefly using prompt by Youbot.
Image Promots using google bard
bard initially gave me the below error when I asked it to generate “Image prompts”.
That’s not something I’m able to do yet.
-google barf
When I removed the work “image, I was able to get three different sets of options, total of 15 prompts! Below are 5 random prompts using a combination of the above:
- Object + Scene + Mood: A lonely cat sitting on a park bench in the rain.
- People + Event + Colors: A group of friends celebrating a wedding with brightly colored balloons and decorations.
- Scene + Style + Mood: A futuristic cityscape in a cyberpunk style, with dark and neon colors and a tense and suspenseful mood.
- Object + Colors + Style: A red rose in a watercolor painting, with soft and delicate brushstrokes.
- Event + Mood: A miracle happening in a war-torn country, with a hopeful and inspiring mood.
A Lonely Cat sitting on a bench in the rain. Image generated using StarryAI
Some additional tips and prompts for stable diffusion using google bard
Be as specific as possible. The more specific you are, the more likely the AI tool is to generate an image that matches your description.
Use keywords. When describing an object or scene, use keywords that are relevant to the image you want to generate. For example, if you want to generate an image of a cat, you could use keywords like “furry,” “whiskers,” and “meow.”
Use adjectives. Adjectives can help you to add more detail to your prompts. For example, if you want to generate an image of a beautiful sunset, you could use adjectives like “colorful,” “majestic,” and “peaceful.”
Use negative keywords. Negative keywords can help you to exclude certain elements from the image you want to generate. For example, if you want to generate an image of a cityscape without any people, you could use the negative keyword “people.”
Experiment. Don’t be afraid to experiment with different prompts to see what kind of images you can generate. The more you experiment, the better you’ll get at creating prompts that produce the results you want.
You can generate image prompts using AI tools such as google bard, YouBot, chatGPT and Anthropic’s Claude.
Image generated using StarryAI.
Note: You can also consider these variations of the images
- A realistic painting of a cat sitting on a windowsill, looking out at a blue sky and white clouds.
- A cartoony drawing of a group of friends having a picnic in a forest, surrounded by trees and flowers.
- A sad photo of a couple holding hands in front of a wedding venue, the bride in tears.
- A detailed painting of a cityscape at night, the skyscrapers lit up against a dark sky.
- A surreal photo of a person standing in a field of red flowers, the sun setting in the background.
AI generated image using Nightcafe.
This post titled Image Prompts for for Stable Diffusion using google bard was published under the category Images. This post was updated in March 2024.
