8 Different Ways to Generate Art using AI
Text to Image
Text to Image (also called Text to Art) is the simplest way to generate images using AI. You just enter a text prompt and AI will generate an image for you. Playform uses customized stable diffusion text-to-image generator. You can also upload an image to be used as a base image to inform the composition and color template.
Unlimited Text-to-Image Generation for Free. Only pay for what you download.
Stylize Filters
Transform an uploaded image to replicate any style of your choice. Input your own image, select from curated styles or optionally upload your style images. You can use any image as a style. Choose from three different processes: Vivid, Pattern, or Fine and control the influence of the style and composition. Results up to 0.4 Megapixel resolution. Generate Unlimited Stylized Images for Free.
Unlimited Generation for Free. Only pay for what you download.
FaceMix
Find faces you're interested in, mix together to create new ones with ultimate fine control. Mix photorealistic and/or illustrative faces and edit face attributes to create any desired effect. Input random generated images, saved images, or your own uploaded images that transform through the process. Results in a grid of 25 or singularly. Generate Unlimited Face Mixes for Free.
Freeform GAN
Train your own AI based on your own Images. This process uses GANs to generate images that are amalgamations of a single collection. Through training, these resulting images resolve more clearly over time. Input one set of 30+ images. The inputted images are the sole source of generation, and having a larger number of consistent images creates better results. Results generate 64 images per snapshot. Freeform GAN only used your images, no pre-trained AI models, so no copyrighted images are used, only your images. You pay only for the hours of training and then generated unlimited images and videos from your trained models. Typically, training is done in one or two hours.
Creative Morph GAN
Train your own AI based on your own Images to morph between two sets of images
This process generates a sequence of images transitioning between two sets: inspirations and influences. Results initially resemble the Inspirations and converge toward the Influences input set.
The images in-between are exploratory mixtures of the two. Input two sets of 30+ images. The Inspiration set informs shapes and contours of results while the Influence set results will iteratively morph to resemble this set.
Each iteration of training generates one output image per Inspiration image provided.
The AI model is trained from scratch, only your images are used in training, so no copyright issues.
You pay for training hours, typically done in less than an hour. Once trained, you can generated unlimited images from that model.
Style Transfer GAN
Using two input sets, this process generates stylized images that use the compositions of the first set and the style of the second. Input two sets of 30+ images. The Inspiration set informs the contours and shapes of the results. The Style set informs the color, texture, and other stylistic elements. Result images generate one output image per Inspiration image provided.
The AI model is trained from scratch, only your images are used in training, so no copyright issues.
You pay for training hours, typically done in half an hour. Once trained, you can generated unlimited images from that model.
Sketch to Art
Upload your original sketches and styles, or draw and apply your favorite artist or movement. Apply styles of your favorite artists, or upload your own style. Results generate 8 images per generation when using provided styles, or 1 image per generation when using an uploaded style.
Unlimited Generation for Free. Only pay for what you download.
Latent Space Video Generation
With trained Freeform GAN model, you can generate “Latent Space” video clips. There are morphing videos generated by choosing key frames from the generations and the model will generate the intermediate frames.
Unlimited Generations for Free. Only pay to train your model.
