🏖️Styler

Using Styler, you can modify the style of input images according to the prompts you choose.

Let's use this image as an example:

Prompt: Describe the painting style you wish to apply, in this example ;et's try having AI mimic Vincent van Gogh's painting style.

Vincent Willem van Gogh painting, Post-Impressionism, Oil painting

Our prompts include only three key terms: Vincent van Gogh's painting, Post-Impressionism, and Oil painting.

Negative Prompt: Specify negative prompts to avoid unwanted content in the generated image.

worst quality, low quality

(Denoising) Strength: This parameter controls the degree to which elements from the original image are retained. A higher value makes the generated image less faithful to the original, while a lower value makes it more faithful. Extreme values represent complete retention and non-retention.

0.75

Classifier-Free Guidance (Scale): This parameter indicates the strength of the prompts, reflecting how closely the AI should adhere to the provided descriptions. A higher value makes the generated image closer to the prompts.

7.5

Number of Inference Steps: 20

Seed: The seed for generating the image. Using the same seed will result in similar style images.

Randomize: Enable this option to ensure a different seed is used each time an image is generated.

Sampler:

The sampler determines the method used for generating images, choose the one you prefer:

  • K_Eular: Suitable for obtaining relatively simple results.

  • DDIM: Emphasizes speed, fusion, novelty, and delivers good-quality results. Step Range: 20-30.

  • DPMSolverMultistep: Aims for high-quality images without concerning whether the image converges.

Image_Guidance_Scale: The similarity between the generated image and the source image. This parameter becomes effective when image_guidance_scale > 1. A higher Image_Guidance_Scale encourages the generated image to resemble the source image, usually at the cost of lower image quality.

jsonCopy codeImage_Guidance_Scale: 1.3
  • If the value is high, image quality is better, but diversity decreases.

  • If the value is low, image quality is worse, but diversity increases.

Finally, we have generated the following image:

Although the painting style has indeed transformed into Vincent van Gogh's, since we included Vincent van Gogh's full name in our prompts, the face of the person in the image has also morphed into Vincent van Gogh's face.

Last updated