Unsupervised domain adaptation framework for crack segmentation using consistency regularization based on cross-domain image augmentation

Deep learning models have obtained state-of-the-art results for medical image analysis. However, CNNs require a massive amount of labelled data to achieve a high performance. Moreover, many supervised learning approaches assume that the training/source dataset and test/target dataset follow the same probability distribution. Nevertheless, this assumption is hardly satisfied in real-world data and when the models are tested on an unseen domain there is a significant performance degradation. In this work, we present an unsupervised Cross-Modality Adversarial Domain Adaptation (C-MADA) framework for medical image segmentation. C-MADA implements an image-level and feature-level adaptation method in a two-step sequential manner. First, images from the source domain are translated to the target domain through an unpaired image-to-image adversarial translation with cycle-consistency loss. Then, a U-Net network is trained with the mapped source domain images and target domain images in an adversarial manner to learn domain-invariant feature representations and produce segmentations for the target domain. Furthermore, to improve the network’s segmentation performance, information about the shape, texture, and contour of the predicted segmentation is included during the adversarial training. C-MADA is tested on the task of brain MRI segmentation from the crossMoDa Grand Challenge and is ranked within the top 15 submissions of the challenge.

Semantic segmentation of transmission-line point clouds is fundamental to intelligent power inspection and grid asset management, as segmentation accuracy directly influences defect detection and facility assessment tasks. However, transmission-line point clouds collected across different voltage levels often show significant variations in density and geometric structure due to heterogeneous LiDAR sensors and flight configurations. Combined with the high cost of large-scale manual annotation, these factors limit the scalability of existing supervised segmentation methods. To overcome these challenges, we propose a geometry-consistency-guided unsupervised domain adaptation framework tailored for cross-voltage transmission-line point-cloud segmentation. The framework employs KPConvX as the backbone and integrates three progressive components. First, a geometric consistency constraint enhances robustness to spatial variations and enables extraction of structural features invariant across voltage levels. Second, a domain feature alignment module reduces distribution shifts through global feature transformation. Third, a minimum-entropy-based pseudo-label refinement strategy improves the reliability of pseudo-labels during self-training. Experiments on a multi-voltage transmission-line dataset demonstrate the effectiveness of the proposed method. With the KPConvX backbone, the framework achieves 66.1% mean Intersection over Union (mIoU) and 94.3% overall accuracy on the unlabeled 110 kV target domain, exceeding the source-only baseline by 15.6% mIoU and outperforming several state-of-the-art UDA methods. This work provides an efficient, annotation-friendly solution for cross-voltage point-cloud segmentation and offers a promising direction for domain adaptation in complex power-grid environments.

Towards collaborative appearance and semantic adaptation for medical image segmentation

Domain adaptation has great values in unpaired cross-modality image segmentation, where the training images with gold standard segmentation are not available from the target image domain. The aim is to reduce the distribution discrepancy between the source and target domains. Hence, an effective measurement for this discrepancy is critical. In this work, we propose a new metric based on characteristic functions of distributions. This metric, referred to as CF distance, enables explicit domain adaptation, in contrast to the implicit manners minimizing domain discrepancy via adversarial training. Based on this CF distance, we propose an unsupervised domain adaptation framework for cross-modality cardiac segmentation, which consists of image reconstruction and prior distribution matching. We validated the method on two tasks, i.e., the CT-MR cross-modality segmentation and the multi-sequence cardiac MR segmentation. Results showed that the proposed explicit metric was effective in domain adaptation, and the segmentation method delivered promising and superior performance, compared to other state-of-the-art techniques. The data and source code of this work has been released via https://zmiclab.github.io/projects.html.

Rapid and accurate acquisition of landslide inventories is essential for effective disaster relief. Deep learning-based pixel-wise semantic segmentation of remote sensing imagery has greatly advanced in landslide mapping. However, the heavy dependance on extensive annotated labels and sensitivity to domain shifts severely constrain the model performance in unseen domains, leading to poor generalization. To address these limitations, we propose LandsDANet, an innovative unsupervised domain adaptation framework for cross-domain landslide identification. Firstly, adversarial learning is employed to reduce the data distribution discrepancies between the source and target domains, thereby achieving output space alignment. The improved SegFormer serves as the segmentation network, incorporating hierarchical Transformer blocks and an attention mechanism to enhance feature representation capabilities. Secondly, to alleviate inter-domain radiometric discrepancies and attain image-level alignment, a Wallis filter is utilized to perform image style transformation. Considering the class imbalance present in the landslide dataset, a Rare Class Sampling strategy is introduced to mitigate bias towards common classes and strengthen the learning of the rare landslide class. Finally, a contrastive loss is adopted to further optimize and enhance the model’s ability to delineate fine-grained class boundaries. The proposed model is validated on the Potsdam and Vaihingen benchmark datasets, followed by validation in two landslide scenarios induced by earthquakes and rainfall to evaluate its adaptability across different disaster domains. Compared to the source-only model, LandsDANet achieved improvements in IoU of 27.04% and 35.73% in two cross-domain landslide disaster recognition tasks, respectively. This performance not only showcases its outstanding capabilities but also underscores its robust potential to meet the demands for rapid response.

Electroencephalogram (EEG) signals convey abundant physiological and psychological information. Decoding EEG is fundamental for brain-computer interaction and medical rehabilitation. Nevertheless, the nonstationarity and inter-individual variability of EEG signals impedes current models from achieving robust crosssubject generalization without expensive subject-specific recalibration, thereby restricting practical deployment. Unsupervised domain adaptation (UDA) aims to improve generalization by minimizing distribution discrepancies between source and target domains. Recent studies regard different subjects as distinct domains and leverage UDA to facilitate cross-subject EEG classification by learning domain-invariant features through discrepancy minimization or adversarial training. However, these conventional methods are all developed in Euclidean space, which are insufficient to capture the non-linear structures inherent in EEG data. To address this limitation, we propose a Deep Adversarial Neural Network on the Symmetric Positive Definite (SPD) matrix manifold, referred to as SPD-DANN, which facilitates the extraction of subject-invariant features through adversarial learning. Additionally, we design an SPD domain feature alignment loss and an SPD class prototype pair loss to simultaneously promote feature alignment across subjects and enhance feature space discriminability. Extensive experiments on four brain-computer interface (BCI) datasets demonstrate that our method surpasses several state-of-the-art UDA techniques. Furthermore, the proposed loss functions are readily adaptable to broader unsupervised or semi-supervised domain adaptation frameworks.

Previous advances in object tracking mostly reported on favorable illumination circumstances while neglecting performance at nighttime, which significantly impeded the development of related aerial robot applications. This work instead develops a novel unsupervised domain adaptation framework for nighttime aerial tracking (named UDAT). Specifically, a unique object discovery approach is provided to generate training patches from raw nighttime tracking videos. To tackle the domain discrepancy, we employ a Transformer-based bridging layer post to the feature extractor to align image features from both domains. With a Transformer day/night feature discriminator, the day-time tracking model is adversarially trained to track at night. Moreover, we construct a pioneering benchmark namely NAT2021 for unsupervised domain adaptive night-time tracking, which comprises a test set of 180 manually annotated tracking sequences and a train set of over 276k unlabelled nighttime tracking frames. Exhaustive experiments demonstrate the robustness and domain adaptability of the proposed framework in nighttime aerial tracking. The code and benchmark are available at https://github.com/vision4robotics/UDAT.

PolSAR (Polimetric Synthetic Aperture Radar) has been shown to be a powerful source of information. As a result of using up to four measurement channels at the same time, which increases the processing depth, it offers information about the geometrical and physical characteristics of objects. However, operating the PolSAR system to its full imaging potential requires significant computing power. In this study, a framework for fully polarimetric SAR image segmentation is proposed, in which the PolSAR signal is decomposed into four components that represent the eigenvectors of the autocovariance matrix corresponding to signals and clutter. The Unsupervised segmentation framework possesses two main processing levels. First level is the data preprocessing, including mean coherency matrix calculation, speckle reduction and polarimetric feature decomposition. Second level include the initial cluster Centers estimation, and edge-region based algorithm. This is achieved by using the combined H-Alpha and (averaged Intensity) lambda features derived from the target decomposition of the PolSAR data. Finally, k-Means clustering based on the Wishart distribution is used to optimize the iterative clustering by merging the clusters with the minimum Wishart distance. The proposed framework is applied on (Flevoland and San_Francisco Bay). The images are selected to react differently with different polarization. The performance evaluation based on qualitative (Visual) and quantitative assessments. Visual assessment provides an excellent information on clarity and delineation of different classes. It is applicable for applications need an accurate statistical information. Quantitative evaluations provide more accurate results for separating different classes in the images. The proposed algorithm is compared to the traditional Cloude-Pottier classification method. The results demonstrate that the proposed algorithm accuracy reaches (88.6 %) with error (0.114), advances over the traditional Cloude-Pottier method with accuracy (84.6 %) and error (0.154).

Existing unsupervised domain adaptation methods based on adversarial learning have achieved good performance in several medical imaging tasks. However, these methods focus only on global distribution adaptation and ignore distribution constraints at the category level, which would lead to sub-optimal adaptation performance. This paper presents an unsupervised domain adaptation framework based on category-level regularization that regularizes the category distribution from three perspectives. Specifically, for inter-domain category regularization, an adaptive prototype alignment module is proposed to align feature prototypes of the same category in the source and target domains. In addition, for intra-domain category regularization, we tailored a regularization technique for the source and target domains, respectively. In the source domain, a prototype-guided discriminative loss is proposed to learn more discriminative feature representations by enforcing intra-class compactness and inter-class separability, and as a complement to traditional supervised loss. In the target domain, an augmented consistency category regularization loss is proposed to force the model to produce consistent predictions for augmented/unaugmented target images, which encourages semantically similar regions to be given the same label. Extensive experiments on two publicly fundus datasets show that the proposed approach significantly outperforms other state-of-the-art comparison algorithms\(^1\)(Our code is available at https://github.com/fengweie/UDA_CLR.).KeywordsUnsupervised domain adaptationCategory level regularizationFundus image segmentation

LUIE: Learnable physical model-guided underwater image enhancement with bi-directional unsupervised domain adaptation

There is growing interest in research on segmentation for the vestibular schwannoma (VS) and cochlea using high-resolution T2 (hrT2) imaging over contrast-enhanced T1 (ceT1) imaging due to the contrast agent side effects. However, the hrT2 imaging remains a problem of insufficient annotated data, which is fatal for building more robust segmentation models. To address the issue, recent studies have adopted unsupervised domain adaptation approaches that translate ceT1 images to hrT2 images. However, previous studies did not consider the size and visual characteristics of the target objects, such as VS and cochlea, during image translation. Specifically, those works simply performed normalization on the entire image without considering its significant impact on the quality of the translated images. These approaches tend to erase the small target objects, making it difficult to preserve the structure of these objects when generating pseudo-target images. Furthermore, they may also struggle to accurately reflect the unique style of the target objects within the images. Therefore, we propose a target-aware unsupervised domain adaptation framework, designed for translating target objects, each tailored to their unique visual characteristics and size using target-aware normalization. We demonstrate the superiority of the proposed framework on a publicly available challenge dataset. Codes are available at https://github.com/Bokyeong-Kang/TANQ.

Unsupervised domain adaptation (UDA) has emerged as an effective paradigm for reducing the huge manual annotation cost for Person Re-Identification (Re-ID). Many of the recent UDA methods for Re-ID are clustering-based and select all the pseudo-label samples in each iteration for the model training. However, there are many wrong labeled samples that will mislead the model optimization under this circumstance. To solve this problem, we propose a Progressive Unsupervised Domain Adaptation (PUDA) framework for image-based Person Re-ID to reduce the negative effect of wrong pseudo-label samples on the model training process. Specifically, we first pretrain a CNN model on a labeled source dataset, then finetune the model on unlabeled target dataset with the following three steps iteratively: 1) estimating pseudo-labels for all the images in the target dataset with the model trained in the last iteration; 2) extending the training set by adding pseudo-label samples with higher label confidence; 3) updating the CNN model with the expanded training set in a supervised manner. During the iteration process, the number of pseudo-label samples added increased progressively. In particular, a Moderate Initial Selections (MIS) strategy for pseudo-label sampling is also proposed to reduce the negative impacts of random noise features in the early iterations and mislabeled samples in the late iterations on the model. The proposed framework with MIS strategy is validated on the Duke-to-Market, Market-to-Duke unsupervised domain adaptation tasks and achieves improvements of 4.2 points (absolute, i.e., 80.0% vs. 75.8%) and 1.7 points (absolute, i.e., 70.7% vs. 69.0%) in mAP correspondingly.

In medical image segmentation, preparing ground truths (or masks) is not trivial as it requires expert clinicians to manually label regions-of-interest. Cervical cytology image segmentation is no exception. In this paper, we propose an unsupervised segmentation framework for cervical cell and whole slide segmentation uses an ensemble of three clustering algorithms namely, K-means, K-means++ and Mean Shift clustering. The final cluster centers obtained from these algorithms are used to initialize cluster points for Fuzzy C-means clustering algorithm. The proposed method is evaluated on multiple standard datasets: HErlev Pap Smear dataset and SIPaKMeD Pap Smear dataset. We also evaluated on a whole slide image dataset (source: CMATER-JU laboratory) and our results are promising and comparable. Overall, our results on multiple benchmark datasets justify the viability of the proposed framework.

Optimal transport (OT) can be applied to cross-domain adaptation in speaker verification (SV) by converting speakers' probability distributions from source to target domains. However, in scenarios involving over-massive categories (speakers) or difficult samples in discrimination, OT often has difficulty computing effective transports. To address this challenge, we propose an OT-based unsupervised domain adaptation (UDA) framework for SV, OT with a diversified memory bank, called DMB-OT, which ensures the accuracy of transfers by two strategies: (1) It regularizes the solution space of OT, which attempts to plan transformations between audio samples from the same speaker with high confidence; (2) it integrates a dynamic curriculum learning algorithm, preventing OT from calculating transport couplings based on hard-discriminative samples in the early stage of UDA. Experiments under different target domains showed that our unsupervised DMB-OT could significantly improve the performance of OT-based UDA and could even match the performance of the supervised PLDA-based adaptation.

This study proposes a novel self‐training framework for unsupervised domain adaptation in the segmentation of concrete wall cracks using accumulated crack data. The proposed method incorporates Bayesian neural networks for uncertainty estimation of pseudo‐labels, and spatial priors of cracks for screening noisy labels. Experiments demonstrate that the proposed approach achieves significant improvements in F1 score. Comparing the F1 scores, Bayesian DeepLabv3+ and Bayesian U‐Net showed performance improvements of 0.0588 and 0.1501, respectively, after domain adaptation. Furthermore, the integration of Stable Diffusion for few‐shot image generation enhances domain adaptation performance by 0.0332. The proposed framework enables high‐precision crack segmentation with as few as 100 target images, which can be easily obtained at the site, reducing the cost of model deployment in infrastructure maintenance. The study also investigates the optimal number of iterations for domain adaptation based on the uncertainty score, providing insights for practical implementation. The proposed method contributes to the development of efficient and automated structural health monitoring using AI.