Make Images of Yourself in Playground AI/Stable Diffusion (without training or downloads)

9 Jan 202306:38

TLDRThe video tutorial demonstrates how to use Playground AI's Stable Diffusion 1.5 to transform a personal photo by altering the background and clothing. The process begins by uploading an image and adjusting its strength to 100 for a direct likeness. A private session is recommended for privacy. The background is then painted over, with attention to contrasting colors, and unwanted elements are removed. The AI generates new images based on the modified background and a selected prompt. Users can iterate by generating more images until they find one they like. Further customization is possible by adjusting filters and experimenting with different settings. The final image can be enhanced with facial restoration and upscaled by four times for higher resolution. The tutorial emphasizes the importance of selecting elements that align with personal taste and the ability to make numerous iterations to achieve the desired outcome.


  • πŸ–ΌοΈ Drag and drop your photo into the provided box to start the conversion process.
  • πŸ”† Increase the image strength to 100 for a direct likeness without variations.
  • 🎨 Use the paint tool to modify the background, ensuring to cover colors that contrast with your skin tone.
  • πŸ–ŒοΈ Select a new background from the available options or use an existing image as a reference.
  • 🧩 Generate multiple images using the same prompt and removal settings to find a suitable background.
  • 🌟 If you don't like the generated images, keep generating until you find one that resonates with you.
  • πŸ‘— Switch to the 'image to image' mode to change your clothing or appearance to match the desired background.
  • 🎭 Play around with different filters and settings to achieve a cinematic or preferred look.
  • πŸ›‘οΈ You can add creative elements like a 'silver knight armor' to your image using the painting mask tool.
  • πŸ”„ Generate multiple variations to find the best fit for your personal preference.
  • πŸ“ˆ Use the 'space restoration' or 'upscale by four' options to enhance the image quality, but note that they apply to different stages of the image.

Q & A

  • What is the first step in converting a photo using the described method?

    -The first step is to drag and drop the image you want to convert into the designated box.

  • What is the significance of setting the image strength to 100?

    -Setting the image strength to 100 ensures that the generated image will be an exact replica of the original, without any variations.

  • Why is it important to modify the background in certain cases?

    -Modifying the background is important to remove contrasting colors, such as red, which can create artifacts and not blend well with the subject's skin tone.

  • How can one ensure they cover colors that differ greatly from their skin tone while modifying the background?

    -One can use the largest brush to paint over areas that contrast with their skin tone and also use the eraser tool to refine the edges as needed.

  • What does the term 'private session' refer to in the context of the script?

    -A private session likely refers to a mode where the user's data and generated images are kept private and not shared or stored publicly.

  • What is the purpose of selecting a background from an external source?

    -Selecting a background from an external source allows the user to create a new image with a desired background that complements the subject of the original photo.

  • How does one generate multiple images with slight variations?

    -One can generate multiple images with slight variations by using the same prompt and removal settings, and then selecting 'generate four images' to create a set of options.

  • What is the benefit of using the 'image to image' feature?

    -The 'image to image' feature allows users to make changes to the original subject of the image, such as altering clothing or style, while maintaining the overall look and feel of the image.

  • How can one refine the generated images to match their personal preference?

    -Users can refine the generated images by using painting masks to modify specific areas, adjusting prompts, and regenerating until they are satisfied with the result.

  • What is the facial restoration feature used for?

    -The facial restoration feature is used to improve the clarity and detail of the subject's face in the generated image.

  • Why is it not recommended to expect both facial restoration and upscaling in a single operation?

    -Facial restoration and upscaling are separate processes. Facial restoration requires downloading and using a specific image, while upscaling increases the resolution of the current image. They serve different purposes and cannot be combined in one step.

  • What should one do if they are happy with one of the generated images?

    -If satisfied with an image, one can use features like space restoration or upscaling to further enhance the image quality before saving it to their desired location.



🎨 Customizing Digital Portraits with Stable Diffusion 1.5

The video script begins with an introduction to using Stable Diffusion 1.5, a tool for editing digital images. The speaker demonstrates how to upload a photo and adjust the image strength to 100 for maximum effect. They then guide viewers through changing the background and clothing of the person in the photo using the tool's painting feature. The process involves selecting and painting over the background, erasing unwanted areas, and ensuring the colors contrast with the subject's skin tone to avoid artifacts. The speaker also discusses the use of prompts to generate new images and the option to upscale the final image by four times, but clarifies that facial restoration and upscaling are separate processes.


✨ Enhancing Image Quality with Upscaling and Restoration

In the second paragraph, the focus shifts to enhancing the quality of the edited image. The speaker talks about using the space restoration feature to improve the image and the option to upscale the image four times for higher resolution. They caution that upscaling does not apply to facial restoration and that the restored image must be downloaded separately. The speaker also shares a common confusion among users regarding the upscaling process and emphasizes the need to download the facial restored image to apply changes. The paragraph concludes with the speaker's decision to upscale the current image by four times and their intention to save the final version to their desktop.



πŸ’‘Stable Diffusion

Stable Diffusion is a term referring to an AI model used for generating images from textual descriptions. In the context of the video, it is the technology that the presenter uses to modify and create new images without the need for training the model or downloading additional software.

πŸ’‘Image Strength

Image strength is a parameter that determines the intensity or fidelity of the image being generated or modified by the AI. In the video, the presenter sets the image strength to 100, which means the generated image will closely resemble the original.

πŸ’‘Private Session

A private session in the context of the video refers to a mode where the AI's image generation is tailored to a user's specific inputs without being influenced by other users' data. This ensures privacy and customization.

πŸ’‘Background Modification

Background modification is the process of changing the backdrop of an image. The video demonstrates how to select and paint over the background to create a new setting for the subject, which is crucial for creating a desired look or theme.

πŸ’‘Paint and Erase Tools

These are digital tools used within the image editing software to modify specific parts of an image. In the video, the presenter uses the paint tool to cover the background and the erase tool to refine the selection, which is essential for the accuracy of the AI's image generation.


In the context of image generation, artifacts refer to unwanted visual elements or distortions that appear in the generated image. The video mentions avoiding red colors in the background to prevent creating artifacts that could detract from the subject.


A prompt is a text input that guides the AI in generating or modifying an image. The video discusses using a prompt to instruct the AI to generate images with specific characteristics, such as a particular background style.

πŸ’‘Image to Image

Image to image is a process where the AI takes an existing image and transforms it into a new image based on a given prompt or set of instructions. This is showcased in the video when the presenter changes the subject's appearance to match a desired style.


Cinematic, in the context of the video, refers to a style of image that resembles the visual quality and composition of a film. The presenter aims to achieve a cinematic look for the generated image to enhance its aesthetic appeal.


Garments are items of clothing. In the video, the presenter discusses changing the subject's clothing within the generated image to something like 'silver knight armor' to fit a specific theme or visual concept.

πŸ’‘Facial Restoration

Facial restoration is a process where the AI enhances or corrects the facial features in an image. The video mentions using facial restoration to improve the clarity and detail of the subject's face in the generated image.


Upscaling is the process of increasing the resolution of an image. In the video, the presenter upscales the image by four times to enhance its detail and sharpness, which is important for larger displays or printing.


Demonstrating how to convert a photo using Playground AI/Stable Diffusion without training or downloads.

Drag and drop a photo to convert it with image strength set to 100 for a direct likeness.

Using Stable Diffusion 1.5 for image processing with private session enabled.

Adjusting lightness to 100 results in generating an identical image.

Modifying the background by painting over it with a brush, ensuring to cover contrasting colors.

Erasing unwanted red background to avoid artifacts and maintain skin tone consistency.

Recovering parts of the image, such as an ear, using the painting tool.

Selecting a new background from Playground AI and customizing it to fit the desired outcome.

Generating four images with the new background and settings to find the most appealing one.

Iteratively generating images until a satisfactory result is achieved.

Using the 'image to image' feature to change the subject's appearance to match a desired style.

Adjusting the filter for a more cinematic look while keeping the subject's likeness.

Experimenting with different clothing, such as a silver knight armor, using an image for image approach.

Generating multiple images to find the best fit for personal preference.

Utilizing facial restoration to improve the quality of the subject's face in the image.

Upscaling the image by four times for higher resolution, noting that it does not apply to the facial restored version.

Avoiding common confusion between upscaling and facial restoration features.

Saving the final upscaled image to a desired location for future use.