Pseudo Labels Research Articles

The data recovery strategy on machine learning against false data injection attacks in power cyber physical systems

During the transmission of power measurement data through communication networks from remote terminal unit (RTU) to the state estimator in Supervisory Control and Data Acquisition (SCADA), power cyber-physical systems (PCPSs) are more susceptible to cyber-attacks. To mitigate that threat, this paper is concerned with a new data recovery strategy on machine learning against false data injection attacks (FDIAs) in PCPSs. Firstly, in view of the limited resources (such as limited energy) of adversaries and system protections, a sparse target false data injection attack (FDIA) is constructed. Then, the FDIA detection problem is transformed into a tripartite separation problem, and the alternating direction method of multipliers on proximal exchange (ADMM-PE) is adopted to complete the intrusion detection of FDIAs. In addition, with the help of reliable mask information and real incomplete measurement data provided by the FDIA detection, a similar supervised generative adversarial imputation networks (GAIN) is proposed to complete the measurement data recovery after FDIAs. Specifically, the pseudo labels generated by data analysis methods such as k-means clustering and support vector machine (SVM) to improve the accuracy of measurement data recovery. Finally, the experimental results of PCPSs show the effectiveness and superiority of the proposed data recovery strategy against FDIAs.

Human-in-the-loop: Using classifier decision boundary maps to improve pseudo labels

For classification tasks, several strategies aim to tackle the problem of not having sufficient labeled data, usually by automatic labeling or by fully passing this task to a user. Automatic labeling is simple to apply but can fail handling complex situations where human insights may be required to decide the correct labels. Conversely, manual labeling leverages the expertise of specialists but may waste precious effort which could be handled by automatic methods. More specifically, automatic solutions could be improved by combining an active learning loop with manual labeling assisted by visual depictions of a classifier’s behavior. We propose to include the human in the labeling loop by using manual labeling in feature spaces produced by a deep feature annotation (DeepFA) technique. To assist manual labeling, we provide users with visual insights on the classifier’s decision boundaries. Finally, we use the manual and automatically computed labels jointly to retrain the classifier in an active learning (AL) loop scheme. Experiments using a toy and a real-world application dataset show that our proposed combination of manual labeling supported by visualization of decision boundaries and automatic labeling can yield a significant increase in classifier performance with a quite limited user effort.

Semi-supervised lung adenocarcinoma histopathology image classification based on multi-teacher knowledge distillation.

In this study, we propose a semi-supervised learning scheme using a patch-based deep learning framework to tackle the challenge of high-precision classification of seven lung tumor growth patterns, despite having a small amount of labeled data in whole slide images (WSIs). This scheme aims to enhance generalization ability with limited data and reduce dependence on large amounts of labeled data. It effectively addresses the common challenge of high demand for labeled data in medical image analysis.

Approach. To address these challenges, the study employs a semi-supervised learning approach enhanced by a dynamic confidence threshold mechanism. This mechanism adjusts based on the quantity and quality of pseudo labels generated. This dynamic thresholding mechanism helps avoid the imbalance of pseudo-label categories and the low number of pseudo-labels that may result from a higher fixed threshold. Furthermore, the research introduces a multi-teacher knowledge distillation technique. This technique adaptively weights predictions from multiple teacher models to transfer reliable knowledge and safeguard student models from low-quality teacher predictions.

Main results. The framework underwent rigorous training and evaluation using a dataset of 150 WSIs, each representing one of the seven growth patterns. The experimental results demonstrate that the framework is highly accurate in classifying lung tumor growth patterns in histopathology images. Notably, the performance of the framework is comparable to that of fully supervised models and human pathologists. In addition, the framework's evaluation metrics on a publicly available dataset are higher than those of previous studies, indicating good generalizability.

Significance. This research demonstrates that a semi-supervised learning approach can achieve results comparable to fully supervised models and expert pathologists, thus opening new possibilities for efficient and cost-effective medical images analysis. The implementation of dynamic confidence thresholding and multi-teacher knowledge distillation techniques represents a significant advancement in applying deep learning to complex medical image analysis tasks. This advancement could lead to faster and more accurate diagnoses, ultimately improving patient outcomes and fostering the overall progress of healthcare technology.&#xD.

Low-resource multi-granularity academic function recognition based on multiple prompt knowledge

PurposeFine-tuning pre-trained language models (PLMs), e.g. SciBERT, generally require large numbers of annotated data to achieve state-of-the-art performance on a range of NLP tasks in the scientific domain. However, obtaining fine-tuning data for scientific NLP tasks is still challenging and expensive. In this paper, the authors propose the mix prompt tuning (MPT), which is a semi-supervised method aiming to alleviate the dependence on annotated data and improve the performance of multi-granularity academic function recognition tasks.Design/methodology/approachSpecifically, the proposed method provides multi-perspective representations by combining manually designed prompt templates with automatically learned continuous prompt templates to help the given academic function recognition task take full advantage of knowledge in PLMs. Based on these prompt templates and the fine-tuned PLM, a large number of pseudo labels are assigned to the unlabelled examples. Finally, the authors further fine-tune the PLM using the pseudo training set. The authors evaluate the method on three academic function recognition tasks of different granularity including the citation function, the abstract sentence function and the keyword function, with data sets from the computer science domain and the biomedical domain.FindingsExtensive experiments demonstrate the effectiveness of the method and statistically significant improvements against strong baselines. In particular, it achieves an average increase of 5% in Macro-F1 score compared with fine-tuning, and 6% in Macro-F1 score compared with other semi-supervised methods under low-resource settings.Originality/valueIn addition, MPT is a general method that can be easily applied to other low-resource scientific classification tasks.

Deep unsupervised shadow detection with curriculum learning and self-training

Shadow detection is undergoing a rapid and remarkable development along with the wide use of deep neural networks. Benefiting from a large number of training images annotated with strong pixel-level ground-truth masks, current deep shadow detectors have achieved state-of-the-art performance. However, it is expensive and time-consuming to provide the pixel-level ground-truth mask for each training image. Considering that, this paper proposes the first unsupervised deep shadow detection framework, which consists of an initial pseudo label generation (IPG) module, a curriculum learning (CL) module and a self-training (ST) module. The supervision signals used in our learning framework are generated from several existing traditional unsupervised shadow detectors, which usually contain a lot of noisy information. Therefore, each module in our unsupervised framework is dedicated to reduce the adverse influence of noisy information on model training. Specifically, the IPG module combines different traditional unsupervised shadow maps to obtain their complementary shadow information. After obtaining the initial pseudo labels, the CL module and the ST module will be used in conjunction to gradually learn new shadow patterns and update the qualities of pseudo labels simultaneously. Extensive experimental results on various benchmark datasets demonstrate that our deep shadow detector not only outperforms the traditional unsupervised shadow detection methods by a large margin but also achieves comparable results with some recent state-of-the-art fully-supervised deep shadow detection methods.

ACAN: A plug-and-play Adaptive Center-Aligned Network for unsupervised domain adaptation

Domain adaptation is an important topic due to its capability in transferring knowledge from source domain to target domain. However, many existing domain adaptation methods primarily concentrate on aligning the data distributions between the source and target domains, often neglecting discriminative feature learning. As a result, target samples with low confidence are embedded near the decision boundary, where they are susceptible to being misclassified, resulting in negative transfer. To address this problem, a novel Adaptive Center-Aligned Network dubbed ACAN is proposed for unsupervised domain adaptation in this work. The main innovations of ACAN are fourfold. Firstly, it is a plug-and-play module and can be easily incorporated into any domain alignment methods without increasing the model complexity and computational burden. Secondly, in contrast to conventional softmax plus cross-entropy loss, angular margin loss is called to enhance the discrimination power for classifier. Thirdly, entropy regularization is exploited to highlight the probability of potential related class, which renders our learned feature representation far away from the decision boundary. Fourthly, to improve the discriminative capacity of model to the target domain, we propose to align the target domain samples to the corresponding class center via pseudo labels. Incorporating ACAN, the performance of baseline domain alignment methods is significantly improved. Extensive ablation and comparison experiments on four widely adopted databases demonstrate the effectiveness of our ACAN. Code is available at: https://github.com/Cloudfly-Z/ACAN

A dynamics simulation-assisted transfer learning method for track condition diagnosis in urban rail transits

Track defects are gradually emerging with the development of urban rail transits. However, there is rare research implemented to diagnose track conditions in real time. Although some intelligent data-driven methods seem to have the potential to achieve the track condition diagnosis, it’s hard to acquire sufficient labeled data in actual applications. This study proposes a dynamics simulation-assisted transfer learning (TL) method for label-scarce track condition diagnosis. Firstly, a dynamics model of axle-box bearings considering the service environment is established. Based on this model, a large amount of axle-box vibration signals corresponding to healthy/defective track conditions is simulated. Wavelet transform is performed for these signals to characterize their time-frequency energy distribution modes in the format of time-frequency maps, which are considered source-domain data. Similarly, the time-frequency maps of the collected signals during vehicle operation are served as the target-domain data. Subsequently, a sub-domain alignment TL network is constructed to map the data from the source and target domain into a deep feature space. In this network, unlabeled target-domain data are classified to obtain their pseudo labels. Finally, Wasserstein distance measure and multiple domain discriminators are employed to achieve label alignment between two domains for each corresponding category. A feature centroid-driven loss function is applied to further reduce the intra-class variations, ultimately realizing accurate knowledge transfer from simulated signals to collected signals. A two-level sliding window algorithm is designed to detect abnormal axle-box vibration signal parts which are then diagnosed through the well-trained network. The proposed method is validated through a transfer diagnosis experiment using simulated signals and collected signals. This study provides a promising solution to diagnose different track conditions, which is of great significance for ensuring running safety in urban rail transits.

Multi-scale feature correspondence and pseudo label retraining strategy for weakly supervised semantic segmentation

Recently, the performance of semantic segmentation using weakly supervised learning has significantly improved. Weakly supervised semantic segmentation (WSSS) that uses only image-level labels has received widespread attention, it employs Class Activation Maps (CAM) to generate pseudo labels. Compared to traditional use of pixel-level labels, this technique greatly reduces annotation costs by utilizing simpler and more readily available image-level annotations. Besides, due to the local perceptual ability of Convolutional Neural Networks (CNN), the generated CAM cannot activate the entire object area. Researchers have found that this CNN limitation can be compensated for by using Vision Transformer (ViT). However, ViT also introduces an over-smoothing problem. Recent research has made good progress in solving this issue, but when discussing CAM and its related segmentation predictions, it is easy to overlook their intrinsic information and the interrelationships between them. In this paper, we propose a Multi-Scale Feature Correspondence (MSFC) method. Our MSFC can obtain the feature correspondence of CAM and segmentation predictions at different scales, re-extract useful semantic information from them, enhancing the network's learning of feature information and improving the quality of CAM. Moreover, to further improve the segmentation precision, we design a Pseudo Label Retraining Strategy (PLRS). This strategy refines the accuracy in local regions, elevates the quality of pseudo labels, and aims to enhance segmentation precision. Experimental results on the PASCAL VOC 2012 and MS COCO 2014 datasets show that our method achieves impressive performance among end-to-end WSSS methods.

A Segment Anything Model based weakly supervised learning method for crop mapping using Sentinel-2 time series images

Automatic crop mapping is essential for various agricultural applications. Although fully convolutional networks (FCNs) have shown effectiveness in crop mapping, they rely on labor-intensive pixel-level annotations. Weakly supervised semantic segmentation (WSSS) methods offer a solution by enabling pixel-level segmentation with less costly annotations, such as point, bounding box (bbox), and image-level annotations. However, these weak annotations often lack complete object information, leading to reduced accuracy. Moreover, there is limited exploration of WSSS methods in medium-resolution satellite imagery, such as Sentinel-2. Therefore, this study proposes SAMWS, a WSSS method based on the Segment Anything Model (SAM) for crop mapping using Sentinel-2 time series images. Leveraging SAM, our method can generate high-quality pseudo labels and accommodate various weak annotations without extensive adjustments. SAMWS comprises three stages: 1) finetuning SAM with adapters; 2) generating pseudo labels using weak annotations and finetuned SAM; and 3) training a segmentation network using pseudo labels for crop mapping. Experiments conducted on PASTIS and Munich datasets demonstrate the superiority of our approach. After SAM was finetuned with adapters, the F1-scores for parcel segmentation using point and bbox prompts increased by 75.0 % and 14.4 %, respectively, reaching 0.880 and 0.931. Additionally, our method achieves classification results closest to fully supervised learning when using point, bbox, and image-level annotations, with F1-scores of 0.810, 0.817, and 0.779, respectively. Our approach offers valuable insights into leveraging foundation models in the remote sensing domain and contains significant potential for crop monitoring. The relevant code will be made publicly available at https://github.com/Nick0317Sun/SAMWS.

OBCTeacher: Resisting labeled data scarcity in oracle bone character detection by semi-supervised learning

Oracle bone characters (OBCs) are ancient ideographs for divination and memorization, as well as first-hand evidence of ancient Chinese culture. The detection of OBC is the premise of advanced studies and was mainly done by authoritative experts in the past. Deep learning techniques have great potential to facilitate OBC detection, but the high annotation cost of OBC brings the scarcity of labeled data, hindering its application. This paper proposes a novel OBC detection framework called OBCTeacher based on semi-supervised learning (SSL) to resist labeled data scarcity. We first construct a large-scale OBC detection dataset. Through investigation, we find that spatial mismatching and class imbalance problems lead to decreased positive anchors and biased predictions, affecting the quality of pseudo labels and the performance of OBC detection. To mitigate the spatial mismatching problem, we introduce a geometric-priori-based anchor assignment strategy and a heatmap polishing procedure to increase positive anchors and improve the quality of pseudo labels. As for the class imbalance problem, we propose a re-weighting method based on estimated class information and a contrastive anchor loss to achieve prioritized learning on different OBC classes and better class boundaries. We evaluate our method by using only a small portion of labeled data while using the remaining data as unlabeled and all labeled data with extra unlabeled data. The results demonstrate the effectiveness of our method compared with other state-of-the-art methods by superior performance and significant improvements of an average of 11.97 in AP50:95 against the only supervised baseline. In addition, our method achieves comparable performance using only 20% of labeled data to the fully-supervised baseline using 100% of labeled data, demonstrating that our method significantly reduces the dependence on labeled data for OBC detection.

Weakly Supervised Lymph Nodes Segmentation Based on Partial Instance Annotations with Pre-trained Dual-branch Network and Pseudo Label Learning

Assessing the presence of potentially malignant lymph nodes aids in estimating cancer progression, and identifying surrounding benign lymph nodes can assist in determining potential metastatic pathways for cancer. For quantitative analysis, automatic segmentation of lymph nodes is crucial. However, due to the labor-intensive and time-consuming manual annotation process required for a large number of lymph nodes, it is more practical to annotate only a subset of the lymph node instances to reduce annotation costs. In this study, we propose a pre-trained Dual-Branch network with Dynamically Mixed Pseudo label (DBDMP) to learn from partial instance annotations for lymph nodes segmentation. To obtain reliable pseudo labels for lymph nodes that are not annotated, we employ a dual-decoder network to generate different outputs that are then dynamically mixed. We integrate the original weak partial annotations with the mixed pseudo labels to supervise the network. To further leverage the extensive amount of unannotated voxels, we apply a self-supervised pre-training strategy to enhance the model’s feature extraction capability. Experiments on the mediastinal Lymph Node Quantification (LNQ) dataset demonstrate that our method, compared to directly learning from partial instance annotations, significantly improves the Dice Similarity Coefficient (DSC) from 11.04% to 54.10% and reduces the Average Symmetric Surface Distance (ASSD) from 20.83 mm to 8.72 mm. The code is available at <a href='https://github.com/WltyBY/LNQ2023_training_code'>https://github.com/WltyBY/LNQ2023_training_code</a>

Progressive expansion for semi-supervised bi-modal salient object detection

Existing bi-modal salient object detection (SOD) methods primarily rely on fully supervised training strategies that require extensive manual annotation. Undoubtedly, extensive manual annotation is time-consuming and laborious, and the fully supervised strategy is also prone to overfitting on the training set. Therefore, we introduce a semi-supervised learning architecture (SSLA) to alleviate these problems while ensuring detection performance. Considering that the inherent training mode and concise architecture of basic SSLA will limit its ability to effectively explore the learning potential of the model, we further propose two optimization strategies, dynamic adjustment and active expansion. Specifically, we dynamically adjust the supervision scheme for unlabeled samples during training so that the model can continuously utilize the model’s gains (pseudo labels) to supervise and guide the model to further explore the unlabeled samples. Furthermore, the active expansion strategy enables the model to acquire more beneficial supervised information and focuses its attention on difficult-to-segment samples. In summary, an effective progressive expansion network (PENet) architecture for semi-supervised bi-modal SOD is proposed. Extensive experiments indicate that our PENet architecture, while effectively alleviating 90% of annotation burdens, has achieved highly competitive results in RGB-T and RGB-D tasks compared to fully supervised methods. The performance is even more pronounced during cross-dataset testing.

Intelligent Classification and Segmentation of Sandstone Thin Section Image Using a Semi-Supervised Framework and GL-SLIC

This study presents the development and validation of a robust semi-supervised learning framework specifically designed for the automated segmentation and classification of sandstone thin section images from the Yanchang Formation in the Ordos Basin. Traditional geological image analysis methods encounter significant challenges due to the labor-intensive and error-prone nature of manual labeling, compounded by the diversity and complexity of rock thin sections. Our approach addresses these challenges by integrating the GL-SLIC algorithm, which combines Gabor filters and Local Binary Patterns for effective superpixel segmentation, laying the groundwork for advanced component identification. The primary innovation of this research is the semi-supervised learning model that utilizes a limited set of manually labeled samples to generate high-confidence pseudo labels, thereby significantly expanding the training dataset. This methodology effectively tackles the critical challenge of insufficient labeled data in geological image analysis, enhancing the model’s generalization capability from minimal initial input. Our framework improves segmentation accuracy by closely aligning superpixels with the intricate boundaries of mineral grains and pores. Additionally, it achieves substantial improvements in classification accuracy across various rock types, reaching up to 96.3% in testing scenarios. This semi-supervised approach represents a significant advancement in computational geology, providing a scalable and efficient solution for detailed petrographic analysis. It not only enhances the accuracy and efficiency of geological interpretations but also supports broader hydrocarbon exploration efforts.

Heterogeneous domain adaptation via incremental discriminative knowledge consistency

Heterogeneous domain adaptation is a challenging problem in transfer learning since samples from the source and target domains reside in different feature spaces with different feature dimensions. The key problem is how to minimize some gaps (e.g., data distribution mismatch) presented in two heterogeneous domains and produce highly discriminative representations for the target domain. In this paper, we attempt to address these challenges with the proposed incremental discriminative knowledge consistency (IDKC) method, which integrates cross-domain mapping, distribution matching, discriminative knowledge preservation, and domain-specific geometry structure consistency into a unified learning model. Specifically, we attempt to learn a domain-specific projection to project original samples into a common subspace in which the marginal distribution is well aligned and the discriminative knowledge consistency is preserved by leveraging the labeled samples from both domains. Moreover, domain-specific structure consistency is enforced to preserve the data manifold from the original space to the common feature space in each domain. Meanwhile, we further apply pseudo labeling to unlabeled target samples based on the feature correlation and retain pseudo labels with high feature correlation coefficients for the next iterative learning. Our pseudo-labeling strategy expands the number of labeled target samples in each category and thus enforces class-discriminative knowledge consistency to produce more discriminative feature representations for the target domain. Extensive experiments on several standard benchmarks for object recognition, cross-language text classification, and digit classification tasks verify the effectiveness of our method.

Multi-view similarity aggregation and multi-level gap optimization for unsupervised person re-identification

Currently, unsupervised person re-identification (Re-ID) typically directly computes the similarity matrix between samples and then applies clustering algorithms to generate pseudo labels for model training, achieving significant progress to some extent. However, distance matrices calculated from a single perspective cannot effectively handle complex and variable sample noise environments, leading to pseudo labels that are not comprehensive or accurate enough, with considerable noise that can misguide model optimization. To address this issue, this paper proposes a Multi-view Similarity Aggregation and Multi-level Gap Optimization (MSAMGO) to improve the accuracy of pseudo labels before and after clustering. Specifically, Multi-view Similarity Aggregation (MSA) involves integrating similarity matrices calculated from multiple perspectives to generate reliable pseudo labels for clustering, greatly reducing noise interference from the source. Additionally, the Multi-level Gap Optimization (MGO) is introduced to leverage the positive and negative centroids of similarity between cluster-level samples and the correlation between instance-level samples. Together, they ensure the overall quality of pseudo-labels, reduce noise interference, and collectively promote model optimization. Experiments were conducted on several datasets, demonstrating that our method surpasses the majority of existing unsupervised person re-identification to a large extent and exhibits excellent generalization.

Structure-aware sign language recognition with spatial–temporal scene graph

Continuous sign language recognition (CSLR) is essential for the social participation of deaf individuals. The structural information of sign language motion units plays a crucial role in semantic representation. However, most existing CSLR methods treat motion units as a whole appearance in the video sequence, neglecting the exploitation and explanation of structural information in the models. This paper proposes a Structure-Aware Graph Convolutional Neural Network (SA-GNN) model for CSLR. This model constructs a spatial–temporal scene graph, explicitly capturing motion units’ spatial structure and temporal variation. Furthermore, to effectively train the SA-GNN, we propose an adaptive bootstrap strategy that enhances weak supervision using dense pseudo labels. This strategy incorporates a confidence cross-entropy loss to adjust the distribution of pseudo labels adaptively. Extensive experiments validate the effectiveness of our proposed method, achieving competitive results on popular CSLR datasets.

Distilling consistent relations for multi-source domain adaptive person re-identification

Although unsupervised person re-identification has made great progress by employing domain adaptation, existing methods rarely make use of multiple source datasets. These works neglect a practical scenario where annotated data from multiple domains are available. To address this issue, we propose a novel method to exploit the diversity and consistency of multiple sources to improve domain adaptation for person re-identification. In the proposed method, multiple expert models pre-trained in different source domains are adapted to the target domain by self-training with expert-specific clustering-based pseudo labels. To improve the consistency of multiple experts, we transfer knowledge between experts through dual similarity distillation between experts. Finally, to further encourage experts to learn complementary features, we maintain the feature diversity of multiple experts by representation decorrelation, preventing them from being too similar. Extensive experiments conducted on three benchmarks, Market-1501, DukeMTMC-reID, and MSMT17, demonstrate the effectiveness of the proposed method.

Source-free domain adaptation via dynamic pseudo labeling and Self-supervision

Recently, unsupervised domain adaptation (UDA) has attracted extensive interest in relieving the greedy requirement of vanilla deep learning for labeled data. It seeks for a solution to adapt the knowledge from a well-labeled training dataset (source domain) to another unlabeled target dataset (target domain). However, in some practical scenarios, the source domain data is inaccessible for a variety of reasons, and only a model trained on it can be provided, thus deriving a more challenging task, i.e., source-free unsupervised domain adaptation (SFUDA). Some pseudo labeling-based methods have been proposed to solve it by predicting pseudo labels for the unlabeled target domain data. Nevertheless, incorrectly designated pseudo labels will impose an adverse impact on the network adaptation. To alleviate this issue, we propose a dynamic confidence-based pseudo labeling strategy for SFUDA in this paper. Unlike those methods that first rigidly assign pseudo labels to all target domain data and then try to weaken the effect of incorrect pseudo labels in training, we proactively label the target samples with higher confidence in a dynamic manner. To further relieve the impact of incorrect pseudo labels, we harness the collaborative learning to constrain the consistency of the network and impose an additional soft supervision. Besides, we also investigate the possible problem brought by our labeling strategy, i.e., the neglect of wavering samples near the decision boundary, and solve it by injecting the self-supervised learning into our model. Experiments on three UDA benchmark datasets demonstrate the state-of-the-art performance of our proposed method. The code is publicly available at https://github.com/meowpass/DPLS.

A novel Parallel Cooperative Mean-Teacher framework (PCMT) combined with prediction uncertainty guide and class contrastive learning for semi-supervised polyp segmentation

Polyp segmentation technology based on deep learning can quickly and accurately help doctors locate lesions, but its development is limited by pixel-level annotations. The polyp segmentation methods based on semi-supervised learning(SSL) is an effective solution to alleviate annotation pressure. For mainstream semi-supervised polyp segmentation methods, due to the extremely deep depth of the polyp segmentation network and the limited use of labeled data for training, the polyp segmentation network is prone to feature expression bottlenecks. And it is a deviation direction to refine pseudo-labels using regional mathematical evaluation, rather than using pixel-level pseudo label refinement schemes. To solve the limitation, based on SSL, a novel Parallel Cooperative Mean-Teacher framework (PCMT) is proposed. PCMT has several effective components: (1) using the segmentation network with deep supervision branches,which is trained by the Parallel Cooperative Mean-Teacher framework for solving feature expression bottlenecks, (2) prediction uncertainty guide for providing more reliable pseudo-labels, (3) class contrastive regularization for strengthening the affinity between the foreground and background features. PCMT and state-of-the-art semi-supervised methods have been compared using 10%, 20%, 50%, and 100% labeled data on five public datasets, respectively. A large number of experimental results show that PCMT has more prominent prediction performance and better generalization ability. Meanwhile, the segmentation performance of PCMT with only requiring 20% labeled data could reach that of full supervised methods with requiring 100% labeled data. Above the results prove that it is feasible of PCMT for semi-supervised polyp segmentation tasks.

Diabetic retinopathy identification based on multi-source-free domain adaptation.

To address the challenges of data labeling difficulties, data privacy, and necessary large amount of labeled data for deep learning methods in diabetic retinopathy (DR) identification, the aim of this study is to develop a source-free domain adaptation (SFDA) method for efficient and effective DR identification from unlabeled data. A multi-SFDA method was proposed for DR identification. This method integrates multiple source models, which are trained from the same source domain, to generate synthetic pseudo labels for the unlabeled target domain. Besides, a softmax-consistence minimization term is utilized to minimize the intra-class distances between the source and target domains and maximize the inter-class distances. Validation is performed using three color fundus photograph datasets (APTOS2019, DDR, and EyePACS). The proposed model was evaluated and provided promising results with respectively 0.8917 and 0.9795 F1-scores on referable and normal/abnormal DR identification tasks. It demonstrated effective DR identification through minimizing intra-class distances and maximizing inter-class distances between source and target domains. The multi-SFDA method provides an effective approach to overcome the challenges in DR identification. The method not only addresses difficulties in data labeling and privacy issues, but also reduces the need for large amounts of labeled data required by deep learning methods, making it a practical tool for early detection and preservation of vision in diabetic patients.