Leonardo AI: Negative Prompts Explained

AI in 5 Minutes
29 Aug 202303:55

TLDRIn this video, the presenter explains how to use Leonardo AI's negative prompt feature to generate high-quality images. The process begins by visiting the Leonardo AI website and signing in with a Google, Apple, or Microsoft account. Once logged in, users navigate to the AI image generation tool and input their desired image prompt, such as a person with specific hair color and style. To refine the image generation, the 'add negative prompt' option is utilized, allowing users to specify elements they wish to avoid in the final image, like multiple heads or distorted features. The video demonstrates how to enhance prompts with advanced specifications, such as ensuring the character has exactly five fingers. After setting the prompts, the image is generated, resulting in a close-to-perfect image that adheres to the user's specifications. The presenter also mentions the 'photo real' feature for more realistic images, which requires more tokens to generate. The video concludes with a thank you note to the viewers.

Takeaways

  • πŸ–ΌοΈ To generate images with Leonardo AI, start by visiting their website and using the AI image generation tool.
  • πŸ”‘ If you have an account, select the 'Yes I'm white listed' option to proceed to the sign-up page.
  • πŸ“‘ You can log in using your Google, Apple, or Microsoft account.
  • πŸ“ Write a normal prompt describing the image you want to generate, such as a person with specific hair color and attire.
  • 🎨 Specify the style of the image, for example, 'Cinematic photography'.
  • βž• Add a negative prompt to exclude unwanted elements from the image, like extra heads or distorted features.
  • πŸ‘οΈβ€πŸ—¨οΈ Use the 'Leonardo diffusion' feature to enhance the image generation process.
  • 🚫 Specify advanced negative prompts if needed, such as avoiding more than five fingers or non-photorealistic elements.
  • ⏱️ After setting your prompts, wait for the AI to create and generate the results.
  • πŸ“š Review the generated images to ensure they meet the criteria set by your prompts, checking for details like the number of fingers.
  • 🐸 For different subjects, keep the prompt consistent and observe how the AI adapts to the negative prompts.
  • πŸ“ˆ For more realistic images, use the 'photo real' feature, which may require more tokens for generation.

Q & A

  • How can one start using Leonardo AI to generate images?

    -To start using Leonardo AI, visit their website, click on 'start using Leonardo', and if you have an account, select 'yes I'm white listed'. You'll be directed to the sign-up page where you can log in with your Google, Apple, or Microsoft account.

  • What is the first step after logging into Leonardo AI?

    -The first step after logging in is to navigate to the AI image generation tool, which is located on the mid-left of the page.

  • How do you create a basic prompt for image generation?

    -You create a basic prompt by typing a description of the image you want to generate, such as 'person with long brown hair, wearing a suit, full body shot'. You can also specify the style, like 'cinematic photography'.

  • What is a negative prompt and how is it used in image generation?

    -A negative prompt is a feature that allows you to specify what you do not want to appear in the generated image. You add a negative prompt by typing out elements you wish to avoid, such as 'two heads', 'weird teeth', or 'distorted hands'.

  • How can you ensure the generated image has more than five fingers on the hands?

    -You can specify 'more than five fingers' in the negative prompt section to ensure that the generated image does not include any hand with fewer than five fingers.

  • What happens after you add both the normal and negative prompts?

    -After adding both prompts, you hit the 'generate' button, and Leonardo AI will create the image based on your specifications. You then wait for the results.

  • How can you check the quality of the generated image?

    -You can check the quality by opening the generated image and looking for the elements you specified in the prompts, ensuring that unwanted features are absent and the desired features are present and accurately depicted.

  • What is the photo-real feature in Leonardo AI?

    -The photo-real feature in Leonardo AI allows for the generation of more realistic images. However, it requires more tokens to generate the image.

  • What should you do if you want to generate another image with a different subject?

    -You can keep the same prompt and just change the subject you want to generate, then hit 'generate' again to create a new image with the specified subject.

  • How does Leonardo AI ensure the generated images are close to perfect?

    -Leonardo AI uses the prompts and negative prompts provided by the user to generate images that closely match the desired outcome, avoiding unwanted elements and focusing on the specified details.

  • What is the significance of the negative prompt feature in Leonardo AI?

    -The negative prompt feature is significant because it gives users the ability to control what they do not want in the generated images, leading to a more refined and personalized image generation process.

  • Can the negative prompt feature be used to enhance the normal prompt?

    -Yes, the negative prompt feature can be used to enhance the normal prompt by specifying unwanted elements, thus allowing for a more precise and higher quality image generation.

Outlines

00:00

🎨 Generating Images with Leonardo's Negative Prompt Feature

The video provides a step-by-step guide on how to use Leonardo AI's image generation tool with negative prompts to create high-quality images. It starts with accessing the Leonardo AI website and signing up or logging in with various account options. The process involves using the AI image generation tool, crafting a normal prompt describing the desired image, and then adding a negative prompt to specify what should be avoided in the generated image. The negative prompt can include elements like 'two heads', 'distorted hands', or 'low quality images'. The video demonstrates the process by generating an image of a person with five fingers and a proper facial structure, and then by generating an image of a frog, both adhering to the provided prompts and avoiding the negative aspects.

Mindmap

Keywords

πŸ’‘Negative Prompts

Negative prompts are instructions given to an AI image generation tool to exclude certain elements from the generated images. In the context of the video, they are used to guide the Leonardo AI in creating images without unwanted features such as 'two heads' or 'distorted hands'. This helps refine the image generation process to meet the user's specific requirements.

πŸ’‘Leonardo AI

Leonardo AI is an AI image generation tool that is the focus of the video. It allows users to generate images based on textual prompts and has a unique feature for negative prompts to refine the image generation process. The tool is accessible through a website where users can log in and start creating images using their desired prompts.

πŸ’‘Image Generation

Image generation refers to the process of creating visual content using artificial intelligence. In the video, Leonardo AI is used to generate images based on the prompts provided by the user. The process involves both positive prompts, which describe the desired elements, and negative prompts, which specify what to avoid in the images.

πŸ’‘AI Image Generation Tool

An AI image generation tool is a software application that uses artificial intelligence to create images from textual descriptions. The tool mentioned in the video, Leonardo AI, is an example of such a tool. It enables users to input prompts and generate images that match those descriptions, with the ability to add negative prompts for more control over the output.

πŸ’‘Cinematic Photography Style

Cinematic photography style refers to the visual aesthetics and techniques used in filmmaking that are applied to still photography to create a dynamic and engaging look. In the video, the user specifies this style when generating an image of a person wearing a suit to give it a dramatic and professional appearance.

πŸ’‘Full Body Shot

A full body shot is a type of photograph or image that captures a person from head to toe. It is mentioned in the video as a requirement for the image of the person wearing a suit, indicating that the user wants a comprehensive view of the subject in the generated image.

πŸ’‘White Listed

Being 'white listed' typically means being granted special access or approval, often in the context of online services. In the video, users who are white listed can bypass certain steps and gain immediate access to the Leonardo AI's image generation features.

πŸ’‘Google, Apple, or Microsoft Account

These are popular platforms for user authentication. In the context of the video, they are mentioned as options for users to log in to the Leonardo AI website, allowing for a seamless sign-in process using credentials from these widely recognized services.

πŸ’‘Tokens

In the context of AI image generation, tokens often refer to the computational resources or units of processing required to generate an image. The video mentions that generating more realistic images with the 'photo real feature' requires more tokens, indicating a higher computational demand for higher quality outputs.

πŸ’‘Photo Real Feature

The photo real feature is an option within the Leonardo AI tool that allows users to generate images that closely resemble real photographs. It is mentioned in the video as a way to enhance the quality of the generated images, albeit at the cost of using more tokens.

πŸ’‘Fingers

In the video, the number of fingers is used as an example of a specific detail that can be controlled through negative prompts. The user specifies that they do not want more than five fingers in the generated image, demonstrating how the tool can be used to correct or avoid common image generation errors.

πŸ’‘Distortions

Distortions refer to any unnatural or unwanted alterations in the appearance of an image. The video emphasizes the use of negative prompts to avoid distortions such as 'non-tometrical face' or 'ugly blur', ensuring that the generated images are free from such imperfections.

Highlights

Introduction to generating images using Leonardo AI's negative prompt feature.

Accessing the Leonardo AI website and starting the image generation process.

Whitelisting and signing up for an account using Google, Apple, or Microsoft credentials.

Navigating to the AI image generation tool on the website.

Writing a normal prompt for the desired image, such as a person with specific hair color and attire.

Specifying the style of the image, for example, 'Cinematic photography'.

Adding a negative prompt to refine the image generation process.

Using Leonardo diffusion to generate images without undesired elements.

Examples of negative prompts include avoiding 'two heads', 'weird teeth', and 'distorted hands'.

Advanced options allow specifying not to have more than five fingers or other specific exclusions.

The image generation process creates results that are checked for adherence to the prompts.

Images are reviewed for accuracy, such as the correct number of fingers and facial structure.

Demonstration of generating a second image with a frog, maintaining the same prompt and negative prompt rules.

The generated images are of high quality and take into account both positive and negative prompts.

Option to enhance realism with the 'photo real' feature, albeit requiring more tokens.

The video concludes with a summary of how to use the negative prompt feature for image generation.

Thanking the viewers for watching the tutorial on Leonardo AI's image generation tool.

Goodbye message signaling the end of the video.