Engineering posts about Diffusion Models
Curated summaries and key learnings for engineers working with Diffusion Models.
DiffuCoder: Understanding and Improving Masked Diffusion Models for Code Generation
The article presents 'DiffuCoder', a masked diffusion model tailored for code generation, highlighting its advantages over autoregressive models. It investigates the denoising processes of diffusion...
DarkDiff: Advancing Low-Light Raw Enhancement by Retasking Diffusion Models for Camera ISP
The article presents DarkDiff, a novel framework that enhances low-light raw images by leveraging pre-trained generative diffusion models in the context of camera image signal processing (ISP)....
Score Distillation of Flow Matching Models
The article presents a novel approach to score distillation in flow matching models, addressing the limitations of diffusion models in image generation. It demonstrates that flow matching can be...
Continuously Augmented Discrete Diffusion model for Categorical Generative Modeling
The article presents the Continuously Augmented Discrete Diffusion (CADD) model, which enhances traditional discrete diffusion models by incorporating a continuous latent space. This approach...