Search Results for author:
Found 7 papers, 3 papers with code
Improving Stability during Upsampling -- on the Importance of Spatial Context
While during downsampling, aliases and artifacts can be reduced by blurring feature maps, the emergence of fine details is crucial during upsampling.
On the unreasonable vulnerability of transformers for image restoration -- and an easy fix
We consider the recently proposed Restormer model, as well as NAFNet and the "Baseline network" which are both simplified versions of a Restormer.
As large as it gets: Learning infinitely large Filters via Neural Implicit Functions in the Fourier Domain
Motivated by the recent trend towards the usage of larger receptive fields for more context-aware neural networks in vision applications, we aim to investigate how large these receptive fields really need to be.
Fix your downsampling ASAP! Be natively more robust via Aliasing and Spectral Artifact free Pooling
Convolutional neural networks encode images through a sequence of convolutions, normalizations and non-linearities as well as downsampling operations into potentially strong semantic embeddings.
Robust Models are less Over-Confident
Further, our analysis of robust models shows that not only AT but also the model's building blocks (like activation functions and pooling) have a strong influence on the models' prediction confidences.
FrequencyLowCut Pooling -- Plug & Play against Catastrophic Overfitting
Over the last years, Convolutional Neural Networks (CNNs) have been the dominating neural architecture in a wide range of computer vision tasks.
Aliasing coincides with CNNs vulnerability towards adversarial attacks
Many commonly well-performing convolutional neural network models have shown to be susceptible to input data perturbations, indicating a low model robustness.
