AI-ordlista
Den kompletta ordlistan över AI
Conditioning Mechanism
Set of architectures and techniques allowing the integration of external information into the diffusion process to guide generation towards specific results.
Sampling Strategy
Sampling method determining how denoising steps are performed in a conditional diffusion model, influencing generation quality and speed.
Text Encoder
Neural network transforming text descriptions into vector representations used as conditions to guide the diffusion process.
Denoising Diffusion
Iterative process of predicting and removing noise to reconstruct a clear image from a noisy state, guided by specific conditions.
Inpainting Diffusion
Conditional diffusion technique allowing to fill or modify specific regions of an image while maintaining consistency with the surrounding context.
DDPM
Denoising Diffusion Probabilistic Models, fundamental approach using Markov processes to model diffusion and denoising with conditioning.
DDIM
Denoising Diffusion Implicit Models, deterministic variant of DDPM allowing faster sampling with fewer steps while maintaining quality.