9 papers
The Score-Difference Flow for Implicit Generative Modeling
Romann M. Weber
Implicit generative modeling (IGM) aims to produce samples of synthetic data matching the characteristics of a target data distribution. Recent work (e.g. score-matching networks,…
Reviving ConvNeXt for Efficient Convolutional Diffusion Models
Taesung Kwon, Lorenzo Bianchi, Lennart Wittke +5
Recent diffusion models increasingly favor Transformer backbones, motivated by the remarkable scalability of fully attentional architectures. Yet the locality bias, parameter effic…
HiGS: History-Guided Sampling for Plug-and-Play Enhancement of Diffusion Models
Seyedmorteza Sadat, Farnood Salehi, Romann M. Weber
While diffusion models have made remarkable progress in image generation, their outputs can still appear unrealistic and lack fine details, especially when using fewer number of ne…
HiWave: Training-Free High-Resolution Image Generation via Wavelet-Based Diffusion Sampling
Tobias Vontobel, Seyedmorteza Sadat, Farnood Salehi +1
Diffusion models have emerged as the leading approach for image synthesis, demonstrating exceptional photorealism and diversity. However, training diffusion models at high resoluti…
Guidance in the Frequency Domain Enables High-Fidelity Sampling at Low CFG Scales
Seyedmorteza Sadat, Tobias Vontobel, Farnood Salehi +1
Classifier-free guidance (CFG) has become an essential component of modern conditional diffusion models. Although highly effective in practice, the underlying mechanisms by which C…
Eliminating Oversaturation and Artifacts of High Guidance Scales in Diffusion Models
Seyedmorteza Sadat, Otmar Hilliges, Romann M. Weber
Classifier-free guidance (CFG) is crucial for improving both generation quality and alignment between the input condition and final output in diffusion models. While a high guidanc…