Image inpainting aims to reconstruct missing regions in corrupted images with semantically consistent content. While modern methods employ perceptual and style losses to enhance inpainting quality by supervising deep feature representations, two key challenges persist: (i) existing approaches necessitate time-consuming grid searches to determine optimal loss weights, and (ii) heterogeneous auxiliary loss terms are assigned fixed weights, limiting their adaptive contributions. To address these limitations, we propose a framework featuring dynamically weighted auxiliary losses and an automated weight adaptation mechanism. Specifically, we introduce Tunable Perceptual Loss (TPL) and Tunable Style Loss (TSL), which generalize traditional perceptual and style losses by incorporating tunable weights that independently scale distinct loss components according to their auxiliary potential. These are optimized via our Adaptive Weight Adjustment (AWA) algorithm, which dynamically reweights TPL and TSL during training by prioritizing loss terms that maximally improve inpainting performance. Empirical evaluations on public datasets demonstrate that our framework enhances state-of-the-art inpainting performance while eliminating manual weight tuning.

The previous relation-based knowledge distillation methods tend to construct global similarity relationship matrix in a mini-batch while ignoring the knowledge of neighbourhood relationship. In this paper, we propose a new similarity-based relational knowledge distillation method that transfers neighbourhood relationship knowledge by selecting K-nearest neighbours for each sample. Our method consists of two components: Neighbourhood Feature Relationship Distillation and Neighbourhood Logits Relationship Distillation. We perform extensive experiments on CIFAR100 and Tiny ImageNet classification datasets and show that our method outperforms the state-of-the-art knowledge distillation methods. Our code is available at: https://github.com/xinxiaoxiaomeng/NRKD.git.

In typical image inpainting tasks, the locations and shapes of damaged or masked areas are often random and irregular. Vanilla convolutions, commonly employed in learning-based inpainting models, treat all spatial features as valid and share parameters across different regions. This approach can struggle with irregular damage patterns, leading to inpainted results that may suffer from color discrepancies and blurriness. In this paper, we introduce a novel operator known as Semi-Independent Convolution (SIConv) to tackle this challenge. The proposed SIConv, on top of the regular convolution with shared weights, also introduces dynamic terms that assign their own independent weights to each part of the image, and the overall computation is formulated as a shared convolution parameter with an additional term to describe the local structure. Qualitative and quantitative experiments demonstrate that our method outperforms the state-of-the-art, yielding clearer, more coherent, and visually convincing inpainting results.

Unsupervised person re-identification aims at identifying the same person across disparate camera feeds without pre-labeled data. Common approaches employ pseudo-label generation to annotate unlabeled datasets, treating them as if they were accurately labeled. However, these methods often rely on single network pre-trained on datasets with a domain gap or irrelevance, leading to suboptimal feature extraction for the re-identification task. To overcome this, we propose an innovative approach that integrates multiple networks, which can mitigate these drawbacks by capturing a broader range of features. Our strategy involves a multi-view clustering technique that leverages features from several networks to improve label accuracy. Additionally, we introduce adversarial examples coupled with a specific adversarial loss to encourage diverse yet complementary learning across the networks. Our methodology was validated on two extensive re-identification datasets, demonstrating its effectiveness through various network combinations.

The identification of rare haplotypes may greatly expand our knowledge in the genetic architecture of both complex and monogenic traits. To this aim, we developed PERHAPS (Paired-End short Reads-based HAPlotyping from next-generation Sequencing data), a new and simple approach to directly call haplotypes from short-read, paired-end Next Generation Sequencing (NGS) data. To benchmark this method, we considered the APOE classic polymorphism (*1/*2/*3/*4), since it represents one of the best examples of functional polymorphism arising from the haplotype combination of two Single Nucleotide Polymorphisms (SNPs). We leveraged the big Whole Exome Sequencing (WES) and SNP-array data obtained from the multi-ethnic UK BioBank (UKBB, N=48,855). By applying PERHAPS, based on piecing together the paired-end reads according to their FASTQ-labels, we extracted the haplotype data, along with their frequencies and the individual diplotype. Concordance rates between WES directly called diplotypes and the ones generated through statistical pre-phasing and imputation of SNP-array data are extremely high (>99%), either when stratifying the sample by SNP-array genotyping batch or self-reported ethnic group. Hardy-Weinberg Equilibrium tests and the comparison of obtained haplotype frequencies with the ones available from the 1000 Genome Project further supported the reliability of PERHAPS. Notably, we were able to determine the existence of the rare APOE*1 haplotype in two unrelated African subjects from UKBB, supporting its presence at appreciable frequency (approximatively 0.5%) in the African Yoruba population. Despite acknowledging some technical shortcomings, PERHAPS represents a novel and simple approach that will partly overcome the limitations in direct haplotype calling from short read-based sequencing.

Semi-supervised person re-identification (Re-ID) is an extension of the existing popular Re-ID research, which only uses a small portion of labeled data, while the majority of the training samples are unlabeled. This paper approaches the problem by constructing a set of heterogeneous convolutional neural networks (CNNs) fine-tuned by utilizing the labeled training samples, and then propagating the labels to the unlabeled portion for further fine-tuning the overall system in a self-paced manner. In this work, a novel self-paced multi-view clustering is presented to generate pseudo labels for unlabeled training samples, which combines multiple heterogeneous CNNs features to cluster. In our clustering method, we introduce a self-paced regularizer to select reliable samples for fine-tuning each CNNs by minimizing ranking loss and identification loss. Specifically, we select a small portion of unlabeled training data when multiple CNNs are weak. With CNNs become stronger, more and more unlabeled samples are selected. Pseudo label estimation and CNNs training are improved simultaneously, which optimize alternatively until all the unlabeled training samples are selected. In our framework, both the optimization of multiple CNNs training and multi-view clustering on unlabeled training samples are self-paced optimizing procedure. Extensive experiments have been conducted on two large-scale Re-ID datasets to demonstrate the superiority of the proposed method.

In this paper, we propose a level-set based segmentation method for medical images with intensity inhomogeneity. Maximum a Posteriori estimation is adopted to combine image segmentation and bias field correction into a unified framework. Within this framework, both contour prior and bias field prior can be fully used. In order to restrict bias field, we introduce an adaptive regularization. Based on this new adaptive regularization, the bias field is estimated more smooth and the input medical image with intensity inhomogeneity is recovered more clearly. Especially, the estimated bias field of our method introduces less structure information obtained from input image. Experimental results on both synthetic and real images show the advantages of our method in both segmentation and bias field correction accuracies as compared with the state-of-the-art approaches.