Quiz - How Stable Diffusion work 1 - Stable Diffusion Art

Why does Stable Diffusion use the latent space? *

To generate higher quality images

To run faster

To increase the size of the model

What mechanism allows the text prompt to control the image generation? *

VAE

Convolutional network

Denoising

Conditioning

What parameter controls how closely should the prompt be followed? *

Denoising strength

Hires fix

CFG Scale

Control weight

In text-to-image, the image is initially in what space? *

Latent space

Pixel space

Base space

Image space

Which part of the Stable Diffusion model is responsible for text conditioning? *

U-Net

VAE encoder

VAE decoder

Refiner

What does the VAE decoder do? *

Convert the image from the pixel space to the latent space.

Convert the image from the latent space to the pixel space.

Decode the image with the tokenizer.

Embedd the image with the text encoder.

Which of the following statement is correct? *

Stable Diffusion denoise a random image in the latent space conditioned by the text prompt.

Stable Diffusion generates an image in the pixel space conditioned by the text prompt.

Stable Diffusion uses VAE to control the text conditioning process.

Stable Diffusion applies conditioning in the VAE network.

Stable Diffusion - Level 1