March 24, 2024Open Access

Efficient Conditional Diffusion Model with Probability Flow Sampling for Image Super-resolution

View Full Paper

Ask AI

Bookmark

Authors

Discussion

Randomized trial reveals improved image super-resolution quality and reduced time consumption, indicating enhanced efficiency.

Generated images show improved quality and consistency, overcoming issues like color shifting and blurriness.
The method achieves a notable reduction in processing time while enhancing super-resolution over previous methods.
This approach utilizes a continuous-time conditional diffusion model and probability flow sampling for efficiency in generation tasks. Extensive testing across datasets like DIV2K and ImageNet supports these findings, showcasing enhanced performance overall.

Yuan et al. (2024) studied this question.

View Full Paper

Ask AI

Bookmark