Figure 7 from DISGAN: Wavelet-informed Discriminator Guides GAN to MRI Super-resolution with Noise Cleaning | Semantic Scholar (2024)

[PDF] Semantic Reader

A new method for 3D SR based on the GAN framework that uses instance noise to balance the GANS training and an updating feature extractor during the training process and produces highly accurate results.

SRGAN, a generative adversarial network (GAN) for image super-resolution (SR), is presented, to its knowledge, the first framework capable of inferring photo-realistic natural images for 4x upscaling factors and a perceptual loss function which consists of an adversarial loss and a content loss.

The proposed generative adversarial network (GAN) with deep multi-attention modules to learn high-frequency information from low-frequency data and a learning method for training domain-specific classifiers as perceptual loss functions results in an efficient and reliable performance.

This work thoroughly study three key components of SRGAN – network architecture, adversarial loss and perceptual loss, and improves each of them to derive an Enhanced SRGAN (ESRGAN), which achieves consistently better visual quality with more realistic and natural textures than SRGAN.

The focus of this work was on constructing images with spatial resolution higher than can be practically obtained by direct Fourier encoding, and a novel learning approach was developed, which was able to provide an estimate of the spatial gradient prior from the low-resolution inputs for the HR reconstruction.

An efficient super-resolution model based on Generative Adversarial Network is described, which produces synthetic images that simulate MR data at ultra high isotropic resolutions of 0 .

A strong baseline model SwinIR is proposed for image restoration based on the Swin Transformer that outperforms state-of-the-art methods on different tasks by up to 0.14∼0.45dB, while the total number of parameters can be reduced byUp to 67%.

This work proposes a residual in residual (RIR) structure to form very deep network, which consists of several residual groups with long skip connections, and proposes a channel attention mechanism to adaptively rescale channel-wise features by considering interdependencies among channels.

The effectiveness of SR3 in cascaded image generation, where a generative model is chained with super-resolution models to synthesize high-resolution images with competitive FID scores on the class-conditional 256×256 ImageNet generation challenge, is shown.

...

...