ScholarGate
Assistent

Methoden vergelijken

Bekijk de geselecteerde methoden naast elkaar; rijen die verschillen zijn gemarkeerd.

Latente Diffusiemodellen×Swin Transformer×
VakgebiedDeep learningDeep learning
FamilieMachine learningMachine learning
Jaar van ontstaan20222021
GrondleggerRobin RombachZe Liu
TypeNeural network architectureNeural network architecture
Oorspronkelijke bronRombach, R., Blattmann, A., Lorenz, D., Esser, P., & Ommer, B. (2022). High-resolution image synthesis with latent diffusion models. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (pp. 10684-10695). DOI ↗Liu, Z., Lin, Y., Cao, Y., Hu, H., Wei, Y., Zhang, Z., Lin, S., & Guo, B. (2021). Swin Transformer: Hierarchical vision transformer using shifted windows. In Proceedings of the IEEE/CVF International Conference on Computer Vision (pp. 10012-10022). DOI ↗
AliassenLDM, Stable Diffusion, Latent DiffusionSwin, Hierarchical Vision Transformer
Verwant44
SamenvattingLatent Diffusion Models (LDMs) are a generative approach introduced by Rombach et al. in 2022 that performs the diffusion process in a compressed latent space rather than pixel space, enabling efficient high-resolution image synthesis. By compressing images into a low-dimensional latent representation using a variational autoencoder, diffusion becomes computationally tractable while maintaining visual quality.The Swin Transformer is a hierarchical vision transformer introduced by Liu et al. in 2021 that uses shifted window attention to achieve computational efficiency while maintaining strong performance on computer vision tasks. Unlike the original Vision Transformer which applies global self-attention, Swin uses local window-based attention with periodic shifting to balance expressiveness and efficiency.
ScholarGateGegevensset
  1. v1
  2. 1 Bronnen
  3. PUBLISHED
  1. v1
  2. 1 Bronnen
  3. PUBLISHED

Naar zoeken Dia's downloaden

ScholarGateMethoden vergelijken: Latent Diffusion Models · Swin Transformer. Geraadpleegd op 2026-06-19 via https://scholargate.app/nl/compare