How to improve AI images

It doesn't have to be complicated

Image credit: Image Creator from Microsoft Designer/DALL.E and Canva

A few days ago, I came across an op-ed courtesy of Sydney Butler (How-To Geek). His article explores the intricacies involved in refining AI-generated images.

Here are my key takeaways:

• Color filters: Whether through dedicated photo editing applications or social media platforms like Instagram, these filters have the ability to imbue images with vibrancy and allure, while concealing imperfections.

• Cropping: By selectively trimming undesirable elements from the edges of a generated image, its overall composition can be refined. Generating images larger than the intended aspect ratio provides ample space for experimentation, enabling the selection of the most visually pleasing portion while discarding the rest.

• Upscaling: Many AI image generation systems struggle to produce high-resolution outputs suitable for printing or professional applications. This is where the AI upscaler comes into play, a tool capable of transforming low-resolution images into sharp, detailed pieces. While some generators offer built-in upscaling options, others may necessitate external solutions such as BigJPG or Gigapixel AI, each offering unique features and varying degrees of effectiveness.

• Inpainting: Inpainting can be a valuable technique for rectifying problematic areas within an image. By delineating specific regions and instructing the AI to focus solely on those areas during regeneration, flaws such as incorrect details or unwanted artifacts can be swiftly removed. This method also proves invaluable for eliminating AI-generated signatures.