Generative Adversarial Networks and Image Synthesis

Subspace Image Synthesis for Style-Invariant Object Recognition

The paper prunes neural networks to isolate computational subspaces for style-invariant object recognition and tests their perceptual relevance through image synthesis.

PathEdit: Diagnostic-Preserving Counterfactual Medical Image Editing through State-Factorized Generative Intervention

PATHEDIT edits medical images to change diagnoses while preserving other patient characteristics, achieving high target success and minimal off-target changes.

Virtual multiplex staining of the pancreatic islets across type 1 diabetes progression using a Schroedinger bridge

This paper introduces SMILE, a diffusion model for converting H&E stains to multiplex IHC images, showing superior performance in pancreatic islet analysis and breast cancer stain conversion.

Probabilistic Brain MR Image Transformation Using Generative Models

This paper compares probabilistic generative models for MR image transformation, finding NCSN most accurate and DDPM best at detecting out-of-distribution inputs.

Unpaired Image-to-Image Translation for Segmentation and Signal Unmixing

Ui2i is a novel unpaired image-to-image translation model that preserves content fidelity in biomedical tasks like nuclear segmentation and signal unmixing.

GestaltGAN: Synthetic photorealistic portraits of individuals with rare genetic disorders

This paper introduces GestaltGAN, a generative model that creates photorealistic portraits of individuals with rare genetic disorders while preserving privacy.

To FID or not to FID: Applying GANs for MRI Image Generation in HPC

This paper generates brain MRI images using GANs in a distributed HPC setting, achieving FID scores comparable to existing works.

Counterfactual MRI Generation with Denoising Diffusion Models for Interpretable Alzheimer's Disease Effect Detection

The paper uses diffusion models to generate synthetic MRI scans for Alzheimer's disease effect detection, improving classifier performance with limited real data.

Generative adversarial networks in cell microscopy for image augmentation. A systematic review

This systematic review analyzes the use of generative adversarial networks for image augmentation in cell microscopy, highlighting common practices, datasets, and challenges in performance metrics.

High-resolution image reconstruction with latent diffusion models from human brain activity

This paper reconstructs high-resolution images from human brain activity using a latent diffusion model, showing high fidelity without additional training.

Energy Guided Diffusion for Generating Neurally Exciting Images

This paper introduces Energy Guided Diffusion (EGG) to generate neurally exciting images, outperforming gradient ascent in generalization and efficiency for macaque V4 models.

SC-GAN: 3D self-attention conditional GAN with spectral normalization for multi-modal neuroimaging synthesis

The paper introduces SC-GAN, a 3D conditional GAN with self-attention and spectral normalization for multi-modal neuroimaging synthesis, showing improved performance over 2D and 3D baselines.

Unsupervised content-preserving transformation for optical microscopy

This paper introduces an unsupervised image transformation method for optical microscopy that avoids data annotation and achieves high-fidelity image transformations across modalities.

GANai: Standardizing CT Images using Generative Adversarial Network with Alternative Improvement

GANai improves CT image standardization by addressing radiomic feature discrepancies through a novel GAN model with an alternative training strategy.

Generative Adversarial Networks Conditioned on Brain Activity Reconstruct Seen Images

This paper uses GANs conditioned on brain activity to reconstruct images from fMRI data, showing they can recover basic stimulus outlines despite noise and limited data.