Federated Learning (FL) allows multiple clients to collaboratively train models without directly sharing their private data. While various data augmentation techniques have been actively studied in the FL environment, most of these methods share input-level or feature-level data information over communication, posing potential privacy leakage. In response to this challenge, we introduce a federated data augmentation algorithm named FedAvP that shares only the augmentation policies, not the data-related information. For data security and efficient policy search, we interpret the policy loss as a meta update loss in standard FL algorithms and utilize the first-order gradient information to further enhance privacy and reduce communication costs. Moreover, we propose a meta-learning method to search for adaptive personalized policies tailored to heterogeneous clients. Our approach outperforms existing best performing augmentation policy search methods and federated data augmentation methods, in the benchmarks for heterogeneous FL.

Federated Learning (FL) aims to train a global inference model from remotely distributed clients, gaining popularity due to its benefit of improving data privacy. However, traditional FL often faces challenges in practical applications, including model overfitting and divergent local models due to limited and non-IID data among clients. To address these issues, we introduce a novel Bayesian meta-learning approach called meta-variational dropout (MetaVD). MetaVD learns to predict client-dependent dropout rates via a shared hypernetwork, enabling effective model personalization of FL algorithms in limited non-IID data settings. We also emphasize the posterior adaptation view of meta-learning and the posterior aggregation view of Bayesian FL via the conditional dropout posterior. We conducted extensive experiments on various sparse and non-IID FL datasets. MetaVD demonstrated excellent classification accuracy and uncertainty calibration performance, especially for out-of-distribution (OOD) clients. MetaVD compresses the local model parameters needed for each client, mitigating model overfitting and reducing communication costs. Code is available at https://github.com/insujeon/MetaVD.

We propose a new GAN-based unsupervised model for disentangled representation learning. The new model is discovered in an attempt to utilize the Information Bottleneck (IB) framework to the optimization of GAN, thereby named IB-GAN. The architecture of IB-GAN is partially similar to that of InfoGAN but has a critical difference; an intermediate layer of the generator is leveraged to constrain the mutual information between the input and the generated output. The intermediate stochastic layer can serve as a learnable latent distribution that is trained with the generator jointly in an end-to-end fashion. As a result, the generator of IB-GAN can harness the latent space in a disentangled and interpretable manner. With the experiments on dSprites and Color-dSprites dataset, we demonstrate that IB-GAN achieves competitive disentanglement scores to those of state-of-the-art β-VAEs and outperforms InfoGAN. Moreover, the visual quality and the diversity of samples generated by IB-GAN are often better than those by β-VAEs and Info-GAN in terms of FID score on CelebA and 3D Chairs dataset.

Recent advances in computation power have allowed computed tomography (CT) to utilise iterative reconstruction (IR) algorithms. The IR technique can handle noisy data and reconstructs optimal CT images from limited projected images. As of cyclic image processing, IR improves the quality of CT images. This approach requires a minimum number of projections to reconstruct an image; however, decreasing the number of projections to 90 can create artefacts and degrade reconstruction quality. To overcome this limitation, the optical flow technique can compute flow vectors between two consecutive projections to generate projected images between frames. Here, optical flow‐based frame interpolation combined with the ordered subset‐modified iterative technique is proposed to reduce computation time, lower the number of projections, and increase reconstruction quality of CT images. The proposed technique can be used to reconstruct a CT image from 90 projections at 4 degree intervals between projection sequences. This approach produces a much better quality reconstruction compared to that produced by an analytical algorithm, which uses 360 projections. The inclusion of an ordered subset reconstructs CT images quickly by accelerating streaming architecture.

Recent advances in computational power have made it possible to use iterative reconstruction (IR) algorithms in clinics for computed tomographic (CT) imaging. Many researchers prefer IR methods to analytical methods because they reduce radiation, image noise, and artefacts. Simultaneous Iterative Reconstruction Technique (SIRT) reduces the number of views needed for CT reconstruction. However, reconstructed images include ray artefacts that can make diagnosis difficult. This study proposes a modified IR algorithm for fast, high‐quality CT reconstruction. The modified method incorporates geometric non‐linear diffusion in the reconstruction estimate to minimise ray artefacts. This method also converges the algorithms into global minima much faster than other methods, using the minimum number of iterations. To meet the high computational demand of improved IR algorithms, a graphics processing unit was used in this study. The authors expect that the proposed technique can be used to reconstruct high‐quality CT images faster and with minimal iterations.

X‐ray image restoration is a difficult problem aimed at improving image quality for visual inspection. Preservation of edges and relevant image information are important for the diagnosis of diseases by physicians or radiologists. In this research, the authors combined a hybrid median filter (HMF) and total variation (TV) minimisation to restore medical X‐ray images. A standard TV minimisation preserves the edges or boundaries of the reconstructed image but also includes false edges. A modified HMF was employed on a Lagrange multiplier associated with a gradient constraint to remove these false edges during the multiplier update. The TV minimisation unconstrained problem was transformed into a constrained form. An augmented Lagrangian method with a variable penalty term was used to handle the constraints in the form for faster convergence, and the alternative direction method solved the subproblems iteratively. The proposed techniques provide improved convergence speed and restore good quality medical X‐ray images to remove false edges and the staircase effect.