Beyond Sharing Weights for Deep Domain Adaptation

Towards the next generation of species delimitation methods: An overview of machine learning applications

Species delimitation is the process of distinguishing between populations of the same species and distinct species of a particular group of organisms. Various methods exist for inferring species limits, whether based on morphological, molecular, or other types of data. In the case of methods based on DNA sequences, most of them are rooted in the coalescent theory. However, coalescence-based models have limitations, for instance regarding complex evolutionary scenarios and large datasets. In this context, machine learning (ML) can be considered as a promising analytical tool, and provides an effective way to explore dataset structures when species-level divergences are hypothesized. In this review, we examine the use of ML in species delimitation and provide an overview and critical appraisal of existing workflows. We also provide simple explanations on how the main types of ML approaches operate, which should help uninitiated researchers and students interested in the field. Our review suggests that while current ML methods designed to infer species limits are analytically powerful, they also present specific limitations and should not be considered as definitive alternatives to coalescent methods for species delimitation. Future ML enterprises to delimit species should consider the constraints related to the use of simulated data, as in other model-based methods relying on simulations. Conversely, the flexibility of ML algorithms offers a significant advantage by enabling the analysis of diverse data types (e.g., genetic and phenotypic) and handling large datasets effectively. We also propose best practices for the use of ML methods in species delimitation, offering insights into potential future applications. We expect that the proposed guidelines will be useful for enhancing the accessibility, effectiveness, and objectivity of ML in species delimitation.

Unsupervised domain adaptation teacher-student network for retinal vessel segmentation via full-resolution refined model

Retinal blood vessels are the only blood vessels in the human body that can be observed non-invasively. Changes in vessel morphology are closely associated with hypertension, diabetes, cardiovascular disease and other systemic diseases, and computers can help doctors identify these changes by automatically segmenting blood vessels in fundus images. If we train a highly accurate segmentation model on one dataset (source domain) and apply it to another dataset (target domain) with a different data distribution, the segmentation accuracy will drop sharply, which is called the domain shift problem. This paper proposes a novel unsupervised domain adaptation method to address this problem. It uses a teacher-student framework to generate pseudo labels for the target domain image, and trains the student network with a combination of source domain loss and domain adaptation loss; finally, the weights of the teacher network are updated from the exponential moving average of the student network and used for the target domain segmentation. We reconstructed the encoder and decoder of the network into a full-resolution refined model by computing the training loss at multiple semantic levels and multiple label resolutions. We validated our method on two publicly available datasets DRIVE and STARE. From STARE to DRIVE, the accuracy, sensitivity, and specificity are 0.9633, 0.8616,and 0.9733, respectively. From DRIVE to STARE, the accuracy, sensitivity, and specificity are 0.9687, 0.8470, and 0.9785, respectively. Our method outperforms most state-of-the-art unsupervised methods. Compared with domain adaptation methods, our method also has the best F1 score (0.8053) from STARE to DRIVE and a competitive F1 score (0.8001) from DRIVE to STARE.

Two-stage adversarial learning based unsupervised domain adaptation for retinal OCT segmentation

BACKGROUND

Deep learning based optical coherence tomography (OCT) segmentation methods have achieved excellent results, allowing quantitative analysis of large-scale data. However, OCT images are often acquired by different devices or under different imaging protocols, which leads to serious domain shift problem. This in turn results in performance degradation of segmentation models.

PURPOSE

Aiming at the domain shift problem, we propose a two-stage adversarial learning based network (TSANet) that accomplishes unsupervised cross-domain OCT segmentation.

METHODS

In the first stage, a Fourier transform based approach is adopted to reduce image style differences from the image level. Then, adversarial learning networks, including a segmenter and a discriminator, are designed to achieve inter-domain consistency in the segmentation output. In the second stage, pseudo labels of selected unlabeled target domain training data are used to fine-tune the segmenter, which further improves its generalization capability. The proposed method was tested on cross-domain datasets for choroid or retinoschisis segmentation tasks. For choroid segmentation, the model was trained on 400 images and validated on 100 images from the source domain, and then trained on 1320 unlabeled images and tested on 330 images from target domain I, and also trained on 400 unlabeled images and tested on 200 images from target domain II. For retinoschisis segmentation, the model was trained on 1284 images and validated on 312 images from the source domain, and then trained on 1024 unlabeled images and tested on 200 images from the target domain.

RESULTS

The proposed method achieved significantly improved results over that without domain adaptation, with improvement of 8.34%, 55.82% and 3.53% in intersection over union (IoU) respectively for the three test sets. The performance is better than some state-of-the-art domain adaptation methods.

CONCLUSIONS

The proposed TSANet, with image level adaptation, feature level adaptation and pseudo-label based fine-tuning, achieved excellent cross-domain generalization. This alleviates the burden of obtaining additional manual labels when adapting the deep learning model to new OCT data.

Dynamic Weighting Translation Transfer Learning for Imbalanced Medical Image Classification

Medical image diagnosis using deep learning has shown significant promise in clinical medicine. However, it often encounters two major difficulties in real-world applications: (1) domain shift, which invalidates the trained model on new datasets, and (2) class imbalance problems leading to model biases towards majority classes. To address these challenges, this paper proposes a transfer learning solution, named Dynamic Weighting Translation Transfer Learning (DTTL), for imbalanced medical image classification. The approach is grounded in information and entropy theory and comprises three modules: Cross-domain Discriminability Adaptation (CDA), Dynamic Domain Translation (DDT), and Balanced Target Learning (BTL). CDA connects discriminative feature learning between source and target domains using a synthetic discriminability loss and a domain-invariant feature learning loss. The DDT unit develops a dynamic translation process for imbalanced classes between two domains, utilizing a confidence-based selection approach to select the most useful synthesized images to create a pseudo-labeled balanced target domain. Finally, the BTL unit performs supervised learning on the reassembled target set to obtain the final diagnostic model. This paper delves into maximizing the entropy of class distributions, while simultaneously minimizing the cross-entropy between the source and target domains to reduce domain discrepancies. By incorporating entropy concepts into our framework, our method not only significantly enhances medical image classification in practical settings but also innovates the application of entropy and information theory within deep learning and medical image processing realms. Extensive experiments demonstrate that DTTL achieves the best performance compared to existing state-of-the-art methods for imbalanced medical image classification tasks.

BERMAD: batch effect removal for single-cell RNA-seq data using a multi-layer adaptation autoencoder with dual-channel framework

MOTIVATION

Removal of batch effect between multiple datasets from different experimental platforms has become an urgent problem, since single-cell RNA sequencing (scRNA-seq) techniques developed rapidly. Although there have been some methods for this problem, most of them still face the challenge of under-correction or over-correction. Specifically, handling batch effect in highly nonlinear scRNA-seq data requires a more powerful model to address under-correction. In the meantime, some previous methods focus too much on removing difference between batches, which may disturb the biological signal heterogeneity of datasets generated from different experiments, thereby leading to over-correction.

RESULTS

In this article, we propose a novel multi-layer adaptation autoencoder with dual-channel framework to address the under-correction and over-correction problems in batch effect removal, which is called BERMAD and can achieve better results of scRNA-seq data integration and joint analysis. First, we design a multi-layer adaptation architecture to model distribution difference between batches from different feature granularities. The distribution matching on various layers of autoencoder with different feature dimensions can result in more accurate batch correction outcome. Second, we propose a dual-channel framework, where the deep autoencoder processing each single dataset is independently trained. Hence, the heterogeneous information that is not shared between different batches can be retained more completely, which can alleviate over-correction. Comprehensive experiments on multiple scRNA-seq datasets demonstrate the effectiveness and superiority of our method over the state-of-the-art methods.

AVAILABILITY AND IMPLEMENTATION

The code implemented in Python and the data used for experiments have been released on GitHub (https://github.com/zhanglabNKU/BERMAD) and Zenodo (https://zenodo.org/records/10695073) with detailed instructions.

Federated Domain Adaptation via Transformer for Multi-Site Alzheimer's Disease Diagnosis

In multi-site studies of Alzheimer's disease (AD), the difference of data in multi-site datasets leads to the degraded performance of models in the target sites. The traditional domain adaptation method requires sharing data from both source and target domains, which will lead to data privacy issue. To solve it, federated learning is adopted as it can allow models to be trained with multi-site data in a privacy-protected manner. In this paper, we propose a multi-site federated domain adaptation framework via Transformer (FedDAvT), which not only protects data privacy, but also eliminates data heterogeneity. The Transformer network is used as the backbone network to extract the correlation between the multi-template region of interest features, which can capture the brain abundant information. The self-attention maps in the source and target domains are aligned by applying mean squared error for subdomain adaptation. Finally, we evaluate our method on the multi-site databases based on three AD datasets. The experimental results show that the proposed FedDAvT is quite effective, achieving accuracy rates of 88.75%, 69.51%, and 69.88% on the AD vs. NC, MCI vs. NC, and AD vs. MCI two-way classification tasks, respectively.

Domain-adaptive neural networks improve supervised machine learning based on simulated population genetic data

Investigators have recently introduced powerful methods for population genetic inference that rely on supervised machine learning from simulated data. Despite their performance advantages, these methods can fail when the simulated training data does not adequately resemble data from the real world. Here, we show that this "simulation mis-specification" problem can be framed as a "domain adaptation" problem, where a model learned from one data distribution is applied to a dataset drawn from a different distribution. By applying an established domain-adaptation technique based on a gradient reversal layer (GRL), originally introduced for image classification, we show that the effects of simulation mis-specification can be substantially mitigated. We focus our analysis on two state-of-the-art deep-learning population genetic methods-SIA, which infers positive selection from features of the ancestral recombination graph (ARG), and ReLERNN, which infers recombination rates from genotype matrices. In the case of SIA, the domain adaptive framework also compensates for ARG inference error. Using the domain-adaptive SIA (dadaSIA) model, we estimate improved selection coefficients at selected loci in the 1000 Genomes CEU population. We anticipate that domain adaptation will prove to be widely applicable in the growing use of supervised machine learning in population genetics.

MiDA: Membership inference attacks against domain adaptation

Domain adaption has become an effective solution to train neural networks with insufficient training data. In this paper, we investigate the vulnerability of domain adaption that potentially breaches sensitive information about the training dataset. We propose a new membership inference attack against domain adaption models, to infer the membership information of samples from the target domain. By leveraging the background knowledge about an additional source-domain in domain adaptation tasks, our attack can exploit the similar distributions between the target and source domain data to determine if a specific data sample belongs in the training set with high efficiency and accuracy. In particular, the proposed attack can be deployed in a practical scenario where the attacker cannot obtain any details of the model. We conduct extensive evaluations for object and digit recognition tasks. Experimental results show that our method can achieve the attack against domain adaptation models with a high success rate.

Anatomy-guided domain adaptation for 3D in-bed human pose estimation

3D human pose estimation is a key component of clinical monitoring systems. The clinical applicability of deep pose estimation models, however, is limited by their poor generalization under domain shifts along with their need for sufficient labeled training data. As a remedy, we present a novel domain adaptation method, adapting a model from a labeled source to a shifted unlabeled target domain. Our method comprises two complementary adaptation strategies based on prior knowledge about human anatomy. First, we guide the learning process in the target domain by constraining predictions to the space of anatomically plausible poses. To this end, we embed the prior knowledge into an anatomical loss function that penalizes asymmetric limb lengths, implausible bone lengths, and implausible joint angles. Second, we propose to filter pseudo labels for self-training according to their anatomical plausibility and incorporate the concept into the Mean Teacher paradigm. We unify both strategies in a point cloud-based framework applicable to unsupervised and source-free domain adaptation. Evaluation is performed for in-bed pose estimation under two adaptation scenarios, using the public SLP dataset and a newly created dataset. Our method consistently outperforms various state-of-the-art domain adaptation methods, surpasses the baseline model by 31%/66%, and reduces the domain gap by 65%/82%. Source code is available at https://github.com/multimodallearning/da-3dhpe-anatomy.

Domain-adaptive neural networks improve supervised machine learning based on simulated population genetic data

Investigators have recently introduced powerful methods for population genetic inference that rely on supervised machine learning from simulated data. Despite their performance advantages, these methods can fail when the simulated training data does not adequately resemble data from the real world. Here, we show that this "simulation mis-specification" problem can be framed as a "domain adaptation" problem, where a model learned from one data distribution is applied to a dataset drawn from a different distribution. By applying an established domain-adaptation technique based on a gradient reversal layer (GRL), originally introduced for image classification, we show that the effects of simulation mis-specification can be substantially mitigated. We focus our analysis on two state-of-the-art deep-learning population genetic methods-SIA, which infers positive selection from features of the ancestral recombination graph (ARG), and ReLERNN, which infers recombination rates from genotype matrices. In the case of SIA, the domain adaptive framework also compensates for ARG inference error. Using the domain-adaptive SIA (dadaSIA) model, we estimate improved selection coefficients at selected loci in the 1000 Genomes CEU population. We anticipate that domain adaptation will prove to be widely applicable in the growing use of supervised machine learning in population genetics.

Cross-Domain Sentiment Analysis Based on Feature Projection and Multi-Source Attention in IoT

Social media is a real-time social sensor to sense and collect diverse information, which can be combined with sentiment analysis to help IoT sensors provide user-demanded favorable data in smart systems. In the case of insufficient data labels, cross-domain sentiment analysis aims to transfer knowledge from the source domain with rich labels to the target domain that lacks labels. Most domain adaptation sentiment analysis methods achieve transfer learning by reducing the domain differences between the source and target domains, but little attention is paid to the negative transfer problem caused by invalid source domains. To address these problems, this paper proposes a cross-domain sentiment analysis method based on feature projection and multi-source attention (FPMA), which not only alleviates the effect of negative transfer through a multi-source selection strategy but also improves the classification performance in terms of feature representation. Specifically, two feature extractors and a domain discriminator are employed to extract shared and private features through adversarial training. The extracted features are optimized by orthogonal projection to help train the classification in multi-source domains. Finally, each text in the target domain is fed into the trained module. The sentiment tendency is predicted in the weighted form of the attention mechanism based on the classification results from the multi-source domains. The experimental results on two commonly used datasets showed that FPMA outperformed baseline models.

Addressing multi-site functional MRI heterogeneity through dual-expert collaborative learning for brain disease identification

Several studies employ multi-site rs-fMRI data for major depressive disorder (MDD) identification, with a specific site as the to-be-analyzed target domain and other site(s) as the source domain. But they usually suffer from significant inter-site heterogeneity caused by the use of different scanners and/or scanning protocols and fail to build generalizable models that can well adapt to multiple target domains. In this article, we propose a dual-expert fMRI harmonization (DFH) framework for automated MDD diagnosis. Our DFH is designed to simultaneously exploit data from a single labeled source domain/site and two unlabeled target domains for mitigating data distribution differences across domains. Specifically, the DFH consists of a domain-generic student model and two domain-specific teacher/expert models that are jointly trained to perform knowledge distillation through a deep collaborative learning module. A student model with strong generalizability is finally derived, which can be well adapted to unseen target domains and analysis of other brain diseases. To the best of our knowledge, this is among the first attempts to investigate multi-target fMRI harmonization for MDD diagnosis. Comprehensive experiments on 836 subjects with rs-fMRI data from 3 different sites show the superiority of our method. The discriminative brain functional connectivities identified by our method could be regarded as potential biomarkers for fMRI-related MDD diagnosis.

A survey on face presentation attack detection mechanisms: hitherto and future perspectives

The advances in human face recognition (FR) systems have recorded sublime success for automatic and secured authentication in diverse domains. Although the traditional methods have been overshadowed by face recognition counterpart during this progress, computer vision gains rapid traction, and the modern accomplishments address problems with real-world complexity. However, security threats in FR-based systems are a growing concern that offers a new-fangled track to the research community. In particular, recent past has witnessed ample instances of spoofing attacks where imposter breaches security of the system with an artifact of human face to circumvent the sensor module. Therefore, presentation attack detection (PAD) capabilities are instilled in the system for discriminating genuine and fake traits and anticipation of their impact on the overall behavior of the FR-based systems. To scrutinize exhaustively the current state-of-the-art efforts, provide insights, and identify potential research directions on face PAD mechanisms, this systematic study presents a review of face anti-spoofing techniques that use computational approaches. The study includes advancements in face PAD mechanisms ranging from traditional hardware-based solutions to up-to-date handcrafted features or deep learning-based approaches. We also present an analytical overview of face artifacts, performance protocols, and benchmark face anti-spoofing datasets. In addition, we perform analysis of the twelve recent state-of-the-art (SOTA) face PAD techniques on a common platform using identical dataset (i.e., REPLAY-ATTACK) and performance protocols (i.e., HTER and ACA). Our overall analysis investigates that despite prevalent face PAD mechanisms demonstrate potential performance, there exist some crucial issues that requisite a futuristic attention. Our analysis put forward a number of open issues such as; limited generalization to unknown attacks, inadequacy of face datasets for DL-models, training models with new fakes, efficient DL-enabled face PAD with smaller datasets, and limited discrimination of handcrafted features. Furthermore, the COVID-19 pandemic is an additional challenge to the existing face-based recognition systems, and hence to the PAD methods. Our motive is to present a complete reference of studies in this field and orient researchers to promising directions.

A novel Pseudo-label based domain adaptation method on tabular data

Tabular data is a widely used data form in many fields such as product marketing. In some cases, the domain shift between source and target domain of tabular data may occur with the changing of collection conditions such as time. The extant methods on tabular data mainly consist of neural-network-based methods and tree-based methods. They both meet challenges induced by domain shift on tabular data. First, neural-network-based methods are lack of effective mechanism to extract the features of tabular data and the performance may not be higher than tree-based models. Second, tree-based methods are lack of effective feature representations to model the associations between source domain and target domain. To improve the performance of tree-based methods for domain shift, a novel pseudo-label based domain adaptation method is proposed for the tree-based method called Xgboost. The proposed method consists of pseudo-label generation and selection strategies. The pseudo-label generation strategy can control the effects of pseudo-labels on Xgboost in a more flexible way by setting proper values of pseudo-labels. The pseudo-label selection strategy can select the pseudo-labels with high confidences under a consistency condition based on the outputs of Xgboost. The quality of pseudo-labels for the data in target domain is improved and so does the performance of Xgboost trained by the data in both source domain and target domain. In the experiment, several UCI datasets and 5G terminal datasets are used to show that the proposed methods can effectively improve the performance of Xgboost.

Improved Domain Adaptation Network Based on Wasserstein Distance for Motor Imagery EEG Classification

Motor Imagery (MI) paradigm is critical in neural rehabilitation and gaming. Advances in brain-computer interface (BCI) technology have facilitated the detection of MI from electroencephalogram (EEG). Previous studies have proposed various EEG-based classification algorithms to identify the MI, however, the performance of prior models was limited due to the cross-subject heterogeneity in EEG data and the shortage of EEG data for training. Therefore, inspired by generative adversarial network (GAN), this study aims to propose an improved domain adaption network based on Wasserstein distance, which utilizes existing labeled data from multiple subjects (source domain) to improve the performance of MI classification on a single subject (target domain). Specifically, our proposed framework consists of three components, including a feature extractor, a domain discriminator, and a classifier. The feature extractor employs an attention mechanism and a variance layer to improve the discrimination of features extracted from different MI classes. Next, the domain discriminator adopts the Wasserstein matrix to measure the distance between source domain and target domain, and aligns the data distributions of source and target domain via adversarial learning strategy. Finally, the classifier uses the knowledge acquired from the source domain to predict the labels in the target domain. The proposed EEG-based MI classification framework was evaluated by two open-source datasets, the BCI Competition IV Datasets 2a and 2b. Our results demonstrated that the proposed framework could enhance the performance of EEG-based MI detection, achieving better classification results compared with several state-of-the-art algorithms. In conclusion, this study is promising in helping the neural rehabilitation of different neuropsychiatric diseases.

Cross-domain few-shot learning based on pseudo-Siamese neural network

Cross-domain few-shot learning is one of the research highlights in machine learning. The difficulty lies in the accuracy drop of cross-domain network learning on a single domain due to the differences between the domains. To alleviate the problem, according to the idea of contour cognition and the process of human recognition, we propose a few-shot learning method based on pseudo-Siamese convolution neural network. The original image and the sketch map are respectively sent to the branch network in the pre-training and meta-learning process. While maintaining the original image features, the contour features are separately extracted as branch for training at the same time to improve the accuracy and generalization of learning. We conduct cross-domain few-shot learning experiments and good results have been achieved using mini-ImageNet as source domain, EuroSAT and ChestX as the target domains. Also, the results are qualitatively analyzed using a heatmap to verify the feasibility of our method.

Domain Adapted Multitask Learning for Segmenting Amoeboid Cells in Microscopy

The method proposed in this paper is a robust combination of multi-task learning and unsupervised domain adaptation for segmenting amoeboid cells in microscopy. A highlight of this work is the manner in which the model's hyperparameters are estimated. The detriments of ad-hoc parameter estimation are well known, but this issue remains largely unaddressed in the context of CNN-based segmentation. Using a novel min-max formulation of the segmentation cost function our proposed method analytically estimates the model's hyperparameters, while simultaneously learning the CNN weights during training. This end-to-end framework provides a consolidated mechanism to harness the potential of multi-task learning to isolate and segment clustered cells from low contrast brightfield images, and it simultaneously leverages deep domain adaptation to segment fluorescent cells without explicit pixel-level re- annotation of the data. Experimental validations on multi-cellular images strongly suggest the effectiveness of the proposed technique, and our quantitative results show at least 15% and 10% improvement in cell segmentation on brightfield and fluorescence images respectively compared to contemporary supervised segmentation methods.

Tackling imbalanced data in cybersecurity with transfer learning: a case with ROP payload detection

In recent years, deep learning gained proliferating popularity in the cybersecurity application domain, since when being compared to traditional machine learning methods, it usually involves less human efforts, produces better results, and provides better generalizability. However, the imbalanced data issue is very common in cybersecurity, which can substantially deteriorate the performance of the deep learning models. This paper introduces a transfer learning based method to tackle the imbalanced data issue in cybersecurity using return-oriented programming payload detection as a case study. We achieved 0.0290 average false positive rate, 0.9705 average F1 score and 0.9521 average detection rate on 3 different target domain programs using 2 different source domain programs, with 0 benign training data sample in the target domain. The performance improvement compared to the baseline is a trade-off between false positive rate and detection rate. Using our approach, the total number of false positives is reduced by 23.16%, and as a trade-off, the number of detected malicious samples decreases by 0.68%.

Contrastive domain adaptation with consistency match for automated pneumonia diagnosis

Pneumonia can be difficult to diagnose since its symptoms are too variable, and the radiographic signs are often very similar to those seen in other illnesses such as a cold or influenza. Deep neural networks have shown promising performance in automated pneumonia diagnosis using chest X-ray radiography, allowing mass screening and early intervention to reduce the severe cases and death toll. However, they usually require many well-labelled chest X-ray images for training to achieve high diagnostic accuracy. To reduce the need for training data and annotation resources, we propose a novel method called Contrastive Domain Adaptation with Consistency Match (CDACM). It transfers the knowledge from different but relevant datasets to the unlabelled small-size target dataset and improves the semantic quality of the learnt representations. Specifically, we design a conditional domain adversarial network to exploit discriminative information conveyed in the predictions to mitigate the domain gap between the source and target datasets. Furthermore, due to the small scale of the target dataset, we construct a feature cloud for each target sample and leverage contrastive learning to extract more discriminative features. Lastly, we propose adaptive feature cloud expansion to push the decision boundary to a low-density area. Unlike most existing transfer learning methods that aim only to mitigate the domain gap, our method instead simultaneously considers the domain gap and the data deficiency problem of the target dataset. The conditional domain adaptation and the feature cloud generation of our method are learning jointly to extract discriminative features in an end-to-end manner. Besides, the adaptive feature cloud expansion improves the model's generalisation ability in the target domain. Extensive experiments on pneumonia and COVID-19 diagnosis tasks demonstrate that our method outperforms several state-of-the-art unsupervised domain adaptation approaches, which verifies the effectiveness of CDACM for automated pneumonia diagnosis using chest X-ray imaging.

Few-Electrode EEG from the Wearable Devices Using Domain Adaptation for Depression Detection

Nowadays, major depressive disorder (MDD) has become a crucial mental disease that endangers human health. Good results have been achieved by electroencephalogram (EEG) signals in the detection of depression. However, EEG signals are time-varying, and the distributions of the different subjects' data are non-uniform, which poses a bad influence on depression detection. In this paper, the deep learning method with domain adaptation is applied to detect depression based on EEG signals. Firstly, the EEG signals are preprocessed and then transformed into pictures by two methods: the first one is to present the three channels of EEG separately in the same image, and the second one is the RGB synthesis of the three channels of EEG. Finally, the training and prediction are performed in the domain adaptation model. The results indicate that the domain adaptation model can effectively extract EEG features and obtain an average accuracy of 77.0 ± 9.7%. This paper proves that the domain adaptation method can effectively weaken the inherent differences of EEG signals, making the diagnosis of different users more accurate.

Attention Regularized Laplace Graph for Domain Adaptation

In leveraging manifold learning in domain adaptation (DA), graph embedding-based DA methods have shown their effectiveness in preserving data manifold through the Laplace graph. However, current graph embedding DA methods suffer from two issues: 1). they are only concerned with preservation of the underlying data structures in the embedding and ignore sub-domain adaptation, which requires taking into account intra-class similarity and inter-class dissimilarity, thereby leading to negative transfer; 2). manifold learning is proposed across different feature/label spaces separately, thereby hindering unified comprehensive manifold learning. In this paper, starting from our previous DGA-DA, we propose a novel DA method, namely A ttention R egularized Laplace G raph-based D omain A daptation (ARG-DA), to remedy the aforementioned issues. Specifically, by weighting the importance across different sub-domain adaptation tasks, we propose the A ttention R egularized Laplace Graph for class aware DA, thereby generating the attention regularized DA. Furthermore, using a specifically designed FEEL strategy, our approach dynamically unifies alignment of the manifold structures across different feature/label spaces, thus leading to comprehensive manifold learning. Comprehensive experiments are carried out to verify the effectiveness of the proposed DA method, which consistently outperforms the state of the art DA methods on 7 standard DA benchmarks, i.e., 37 cross-domain image classification tasks including object, face, and digit images. An in-depth analysis of the proposed DA method is also discussed, including sensitivity, convergence, and robustness.

Generalized Meta-FDMixup: Cross-Domain Few-Shot Learning Guided by Labeled Target Data

The vanilla Few-shot Learning (FSL) learns to build a classifier for a new concept from one or very few target examples, with the general assumption that source and target classes are sampled from the same domain. Recently, the task of Cross-Domain Few-Shot Learning (CD-FSL) aims at tackling the FSL where there is a huge domain shift between the source and target datasets. Extensive efforts on CD-FSL have been made via either directly extending the meta-learning paradigm of vanilla FSL methods, or employing massive unlabeled target data to help learn models. In this paper, we notice that in the CD-FSL task, the few labeled target images have never been explicitly leveraged to inform the model in the training stage. However, such a labeled target example set is very important to bridge the huge domain gap. Critically, this paper advocates a more practical training scenario for CD-FSL. And our key insight is to utilize a few labeled target data to guide the learning of the CD-FSL model. Technically, we propose a novel Generalized Meta-learning based Feature-Disentangled Mixup network, namely GMeta-FDMixup. We make three key contributions of utilizing GMeta-FDMixup to address CD-FSL. Firstly, we present two mixup modules - mixup-P and mixup-M that help facilitate utilizing the unbalanced and disjoint source and target datasets. These two novel modules enable diverse image generation for training the model on the source domain. Secondly, to narrow the domain gap explicitly, we contribute a novel feature disentanglement module that learns to decouple the domain-irrelevant and domain-specific features. By stripping the domain-specific features, we alleviate the negative effects caused by the domain inductive bias. Finally, we repurpose a new contrastive learning module, dubbed ConL. ConL prevents the model from only capturing category-related features via introducing contrastive loss. Thus, the generalization ability on novel categories is improved. Extensive experimental results on two benchmarks show the superiority of our setting and the effectiveness of our method. Code and models will be released.

Harmonization Strategies in Multicenter MRI-Based Radiomics

Radiomics analysis is a powerful tool aiming to provide diagnostic and prognostic patient information directly from images that are decoded into handcrafted features, comprising descriptors of shape, size and textural patterns. Although radiomics is gaining momentum since it holds great promise for accelerating digital diagnostics, it is susceptible to bias and variation due to numerous inter-patient factors (e.g., patient age and gender) as well as inter-scanner ones (different protocol acquisition depending on the scanner center). A variety of image and feature based harmonization methods has been developed to compensate for these effects; however, to the best of our knowledge, none of these techniques has been established as the most effective in the analysis pipeline so far. To this end, this review provides an overview of the challenges in optimizing radiomics analysis, and a concise summary of the most relevant harmonization techniques, aiming to provide a thorough guide to the radiomics harmonization process.

Divergence-Agnostic Unsupervised Domain Adaptation by Adversarial Attacks

Conventional machine learning algorithms suffer the problem that the model trained on existing data fails to generalize well to the data sampled from other distributions. To tackle this issue, unsupervised domain adaptation (UDA) transfers the knowledge learned from a well-labeled source domain to a different but related target domain where labeled data is unavailable. The majority of existing UDA methods assume that data from the source domain and the target domain are available and complete during training. Thus, the divergence between the two domains can be formulated and minimized. In this paper, we consider a more practical yet challenging UDA setting where either the source domain data or the target domain data are unknown. Conventional UDA methods would fail this setting since the domain divergence is agnostic due to the absence of the source data or the target data. Technically, we investigate UDA from a novel view-adversarial attack-and tackle the divergence-agnostic adaptive learning problem in a unified framework. Specifically, we first report the motivation of our approach by investigating the inherent relationship between UDA and adversarial attacks. Then we elaborately design adversarial examples to attack the training model and harness these adversarial examples. We argue that the generalization ability of the model would be significantly improved if it can defend against our attack, so as to improve the performance on the target domain. Theoretically, we analyze the generalization bound for our method based on domain adaptation theories. Extensive experimental results on multiple UDA benchmarks under conventional, source-absent and target-absent UDA settings verify that our method is able to achieve a favorable performance compared with previous ones. Notably, this work extends the scope of both domain adaptation and adversarial attack, and expected to inspire more ideas in the community.

Unsupervised domain adaptive myocardial infarction MRI classification diagnostics model based on target domain confused sample resampling

OBJECTIVE

Inefficient circulatory system due to blockage of blood vessels leads to myocardial infarction and acute blockage. Myocardial infarction is frequently classified and diagnosed in medical treatment using MRI, yet this method is ineffective and prone to error. As a result, there are several implementation scenarios and clinical significance for employing deep learning to develop computer-aided algorithms to aid cardiologists in the routine examination of cardiac MRI.

METHODS

This research uses two distinct domain classifiers to address this issue and achieve domain adaptation between the particular field and the specific part is a problem Current research on environment adaptive systems cannot effectively obtain and apply classification information for unsupervised scenes of target domain images. Insufficient information interchange between specific domains and specific domains is a problem. In this study, two different domain classifiers are used to solve this problem and achieve domain adaption. To effectively mine the source domain images for classification understanding, an unsupervised MRI classification technique for myocardial infarction called CardiacCN is proposed, which relies on adversarial instructions related to the interpolation of confusion specimens in the target domain for the conflict of confusion specimens for the target domain classification task.

RESULTS

The experimental results demonstrate that the CardiacCN model in this study performs better on the six domain adaption tasks of the Sunnybrook Cardiac Dataset (SCD) dataset and increases the mean target area myocardial infarction MRI classification accuracy by approximately 1.2 percent. The classification performance of the CardiacCN model on the target domain does not vary noticeably when the temperature-controlled duration hyper-parameter rl falls in the region of 5-30. According to the experimental findings, the CardiacCN model is more resistant to the excitable rl. The CardiacCN model suggested in this research may successfully increase the accuracy of the source domain predictor for the target domain myocardial infarction clinical scanning classification in unsupervised learning, as shown by the visualization analysis infrastructure provision nurture. It is evident from the visualization assessment of embedded features that the CardiacCN model may significantly increase the source domain classifier's accuracy for the target domain's classification of myocardial infarction in clinical scans under unsupervised conditions.

CONCLUSION

To address misleading specimens with the inconsistent classification of target-domain myocardial infarction medical scans, this paper introduces the CardiacCN unsupervised domain adaptive MRI classification model, which relies on adversarial learning associated with resampling target-domain confusion samples. With this technique, implicit image classification information from the target domain is fully utilized, knowledge transfer from the target domain to the specific domain is encouraged, and the classification effect of the myocardial ischemia medical scan is improved in the target domain of the unsupervised scene.

Towards Uncovering the Intrinsic Data Structures for Unsupervised Domain Adaptation Using Structurally Regularized Deep Clustering

Unsupervised domain adaptation (UDA) is to learn classification models that make predictions for unlabeled data on a target domain, given labeled data on a source domain whose distribution diverges from the target one. Mainstream UDA methods strive to learn domain-aligned features such that classifiers trained on the source features can be readily applied to the target ones. Although impressive results have been achieved, these methods have a potential risk of damaging the intrinsic data structures of target discrimination, raising an issue of generalization particularly for UDA tasks in an inductive setting. To address this issue, we are motivated by a UDA assumption of structural similarity across domains, and propose to directly uncover the intrinsic target discrimination via constrained clustering, where we constrain the clustering solutions using structural source regularization that hinges on the very same assumption. Technically, we propose a hybrid model of Structurally Regularized Deep Clustering, which integrates the regularized discriminative clustering of target data with a generative one, and we thus term our method as H-SRDC. Our hybrid model is based on a deep clustering framework that minimizes the Kullback-Leibler divergence between the distribution of network prediction and an auxiliary one, where we impose structural regularization by learning domain-shared classifier and cluster centroids. By enriching the structural similarity assumption, we are able to extend H-SRDC for a pixel-level UDA task of semantic segmentation. We conduct extensive experiments on seven UDA benchmarks of image classification and semantic segmentation. With no explicit feature alignment, our proposed H-SRDC outperforms all the existing methods under both the inductive and transductive settings. We make our implementation codes publicly available at https://github.com/huitangtang/H-SRDC.

LMISA: A lightweight multi-modality image segmentation network via domain adaptation using gradient magnitude and shape constraint

In medical image segmentation, supervised machine learning models trained using one image modality (e.g. computed tomography (CT)) are often prone to failure when applied to another image modality (e.g. magnetic resonance imaging (MRI)) even for the same organ. This is due to the significant intensity variations of different image modalities. In this paper, we propose a novel end-to-end deep neural network to achieve multi-modality image segmentation, where image labels of only one modality (source domain) are available for model training and the image labels for the other modality (target domain) are not available. In our method, a multi-resolution locally normalized gradient magnitude approach is firstly applied to images of both domains for minimizing the intensity discrepancy. Subsequently, a dual task encoder-decoder network including image segmentation and reconstruction is utilized to effectively adapt a segmentation network to the unlabeled target domain. Additionally, a shape constraint is imposed by leveraging adversarial learning. Finally, images from the target domain are segmented, as the network learns a consistent latent feature representation with shape awareness from both domains. We implement both 2D and 3D versions of our method, in which we evaluate CT and MRI images for kidney and cardiac tissue segmentation. For kidney, a public CT dataset (KiTS19, MICCAI 2019) and a local MRI dataset were utilized. The cardiac dataset was from the Multi-Modality Whole Heart Segmentation (MMWHS) challenge 2017. Experimental results reveal that our proposed method achieves significantly higher performance with a much lower model complexity in comparison with other state-of-the-art methods. More importantly, our method is also capable of producing superior segmentation results than other methods for images of an unseen target domain without model retraining. The code is available at GitHub (https://github.com/MinaJf/LMISA) to encourage method comparison and further research.

Contradistinguisher: A Vapnik's Imperative to Unsupervised Domain Adaptation

Recent domain adaptation works rely on an indirect way of first aligning the source and target domain distributions and then train a classifier on the labeled source domain to classify the target domain. However, the main drawback of this approach is that obtaining a near-perfect domain alignment in itself might be difficult/impossible (e.g., language domains). To address this, inspired by how humans use supervised-unsupervised learning to perform tasks seamlessly across multiple domains or tasks, we follow Vapnik's imperative of statistical learning that states any desired problem should be solved in the most direct way rather than solving a more general intermediate task and propose a direct approach to domain adaptation that does not require domain alignment. We propose a model referred to as Contradistinguisher that learns contrastive features and whose objective is to jointly learn to contradistinguish the unlabeled target domain in an unsupervised way and classify in a supervised way on the source domain. We achieve the state-of-the-art on Office-31, Digits and VisDA-2017 datasets in both single-source and multi-source settings. We demonstrate that performing data augmentation results in an improvement in the performance over vanilla approach. We also notice that the contradistinguish-loss enhances performance by increasing the shape bias.

Addressing modern and practical challenges in machine learning: a survey of online federated and transfer learning

Online federated learning (OFL) and online transfer learning (OTL) are two collaborative paradigms for overcoming modern machine learning challenges such as data silos, streaming data, and data security. This survey explores OFL and OTL throughout their major evolutionary routes to enhance understanding of online federated and transfer learning. Practical aspects of popular datasets and cutting-edge applications for online federated and transfer learning are also highlighted in this work. Furthermore, this survey provides insight into potential future research areas and aims to serve as a resource for professionals developing online federated and transfer learning frameworks.

A Survey on Adversarial Domain Adaptation

Having a lot of labeled data is always a problem in machine learning issues. Even by collecting lots of data hardly, shift in data distribution might emerge because of differences in source and target domains. The shift would make the model to face with problems in test step. Therefore, the necessity of using domain adaptation emerges. There are three techniques in the field of domain adaptation namely discrepancy based, adversarial based and reconstruction based methods. For domain adaptation, adversarial learning approaches showed state-of-the-art performance. Although there are some comprehensive surveys about domain adaptation, we technically focus on adversarial based domain adaptation methods. We examine each proposed method in detail with respect to their structures and objective functions. The common aspect of proposed methods besides domain adaptation is considering the target labels are predicted as accurately as possible. It can be represented by some methods such as metric learning and multi-adversarial discriminators as are used in some of the papers. Also, we address the negative transfer issue for dissimilar distributions and propose the addition of clustering heuristics to the underlying structures for future research.

Active Domain Adaptation With Application to Intelligent Logging Lithology Identification

Lithology identification plays an essential role in formation characterization and reservoir exploration. As an emerging technology, intelligent logging lithology identification has received great attention recently, which aims to infer the lithology type through the well-logging curves using machine-learning methods. However, the model trained on the interpreted logging data is not effective in predicting new exploration well due to the data distribution discrepancy. In this article, we aim to train a lithology identification model for the target well using a large amount of source-labeled logging data and a small amount of target-labeled data. The challenges of this task lie in three aspects: 1) the distribution misalignment; 2) the data divergence; and 3) the cost limitation. To solve these challenges, we propose a novel active adaptation for logging lithology identification (AALLI) framework that combines active learning (AL) and domain adaptation (DA). The contributions of this article are three-fold: 1) the domain-discrepancy problem in intelligent logging lithology identification is first investigated in this article, and a novel framework that incorporates AL and DA into lithology identification is proposed to handle the problem; 2) we design a discrepancy-based AL and pseudolabeling (PL) module and an instance importance weighting module to query the most uncertain target information and retain the most confident source information, which solves the challenges of cost limitation and distribution misalignment; and 3) we develop a reliability detecting module to improve the reliability of target pseudolabels, which, together with the discrepancy-based AL and PL module, solves the challenge of data divergence. Extensive experiments on three real-world well-logging datasets demonstrate the effectiveness of the proposed method compared to the baselines.

Dynamic Instance Domain Adaptation

Most existing studies on unsupervised domain adaptation (UDA) assume that each domain's training samples come with domain labels (e.g., painting, photo). Samples from each domain are assumed to follow the same distribution and the domain labels are exploited to learn domain-invariant features via feature alignment. However, such an assumption often does not hold true-there often exist numerous finer-grained domains (e.g., dozens of modern painting styles have been developed, each differing dramatically from those of the classic styles). Therefore, forcing feature distribution alignment across each artificially-defined and coarse-grained domain can be ineffective. In this paper, we address both single-source and multi-source UDA from a completely different perspective, which is to view each instance as a fine domain. Feature alignment across domains is thus redundant. Instead, we propose to perform dynamic instance domain adaptation (DIDA). Concretely, a dynamic neural network with adaptive convolutional kernels is developed to generate instance-adaptive residuals to adapt domain-agnostic deep features to each individual instance. This enables a shared classifier to be applied to both source and target domain data without relying on any domain annotation. Further, instead of imposing intricate feature alignment losses, we adopt a simple semi-supervised learning paradigm using only a cross-entropy loss for both labeled source and pseudo labeled target data. Our model, dubbed DIDA-Net, achieves state-of-the-art performance on several commonly used single-source and multi-source UDA datasets including Digits, Office-Home, DomainNet, Digit-Five, and PACS.

DeepDA-Ace: A Novel Domain Adaptation Method for Species-Specific Acetylation Site Prediction

Protein lysine acetylation is an important type of post-translational modification (PTM), and it plays a crucial role in various cellular processes. Recently, although many researchers have focused on developing tools for acetylation site prediction based on computational methods, most of these tools are based on traditional machine learning algorithms for acetylation site prediction without species specificity, still maintained as a single prediction model. Recent studies have shown that the acetylation sites of distinct species have evident location-specific differences; however, there is currently no integrated prediction model that can effectively predict acetylation sites cross all species. Therefore, to enhance the scope of species-specific level, it is necessary to establish a framework for species-specific acetylation site prediction. In this work, we propose a domain adaptation framework DeepDA-Ace for species-specific acetylation site prediction, including Rattus norvegicus, Schistosoma japonicum, Arabidopsis thaliana, and other types of species. In DeepDA-Ace, an attention based densely connected convolutional neural network is designed to capture sequence features, and the semantic adversarial learning strategy is proposed to align features of different species so as to achieve knowledge transfer. The DeepDA-Ace outperformed both the general prediction model and fine-tuning based species-specific model across most types of species. The experiment results have demonstrated that DeepDA-Ace is superior to the general and fine-tuning methods, and its precision exceeds 0.75 on most species. In addition, our method achieves at least 5% improvement over the existing acetylation prediction tools.

Unsupervised Multi-Class Domain Adaptation: Theory, Algorithms, and Practice

In this paper, we study the formalism of unsupervised multi-class domain adaptation (multi-class UDA), which underlies a few recent algorithms whose learning objectives are only motivated empirically. Multi-Class Scoring Disagreement (MCSD) divergence is presented by aggregating the absolute margin violations in multi-class classification, and this proposed MCSD is able to fully characterize the relations between any pair of multi-class scoring hypotheses. By using MCSD as a measure of domain distance, we develop a new domain adaptation bound for multi-class UDA; its data-dependent, probably approximately correct bound is also developed that naturally suggests adversarial learning objectives to align conditional feature distributions across source and target domains. Consequently, an algorithmic framework of Multi-class Domain-adversarial learning Networks (McDalNets) is developed, and its different instantiations via surrogate learning objectives either coincide with or resemble a few recently popular methods, thus (partially) underscoring their practical effectiveness. Based on our identical theory for multi-class UDA, we also introduce a new algorithm of Domain-Symmetric Networks (SymmNets), which is featured by a novel adversarial strategy of domain confusion and discrimination. SymmNets affords simple extensions that work equally well under the problem settings of either closed set, partial, or open set UDA. We conduct careful empirical studies to compare different algorithms of McDalNets and our newly introduced SymmNets. Experiments verify our theoretical analysis and show the efficacy of our proposed SymmNets. In addition, we have made our implementation code publicly available.

Domain Adaptation Using Convolutional Autoencoder and Gradient Boosting for Adverse Events Prediction in the Intensive Care Unit

More than 5 million patients have admitted annually to intensive care units (ICUs) in the United States. The leading causes of mortality are cardiovascular failures, multi-organ failures, and sepsis. Data-driven techniques have been used in the analysis of patient data to predict adverse events, such as ICU mortality and ICU readmission. These models often make use of temporal or static features from a single ICU database to make predictions on subsequent adverse events. To explore the potential of domain adaptation, we propose a method of data analysis using gradient boosting and convolutional autoencoder (CAE) to predict significant adverse events in the ICU, such as ICU mortality and ICU readmission. We demonstrate our results from a retrospective data analysis using patient records from a publicly available database called Multi-parameter Intelligent Monitoring in Intensive Care-II (MIMIC-II) and a local database from Children's Healthcare of Atlanta (CHOA). We demonstrate that after adopting novel data imputation on patient ICU data, gradient boosting is effective in both the mortality prediction task and the ICU readmission prediction task. In addition, we use gradient boosting to identify top-ranking temporal and non-temporal features in both prediction tasks. We discuss the relationship between these features and the specific prediction task. Lastly, we indicate that CAE might not be effective in feature extraction on one dataset, but domain adaptation with CAE feature extraction across two datasets shows promising results.

Small Data Challenges in Big Data Era: A Survey of Recent Progress on Unsupervised and Semi-Supervised Methods

Representation learning with small labeled data have emerged in many problems, since the success of deep neural networks often relies on the availability of a huge amount of labeled data that is expensive to collect. To address it, many efforts have been made on training sophisticated models with few labeled data in an unsupervised and semi-supervised fashion. In this paper, we will review the recent progresses on these two major categories of methods. A wide spectrum of models will be categorized in a big picture, where we will show how they interplay with each other to motivate explorations of new ideas. We will review the principles of learning the transformation equivariant, disentangled, self-supervised and semi-supervised representations, all of which underpin the foundation of recent progresses. Many implementations of unsupervised and semi-supervised generative models have been developed on the basis of these criteria, greatly expanding the territory of existing autoencoders, generative adversarial nets (GANs) and other deep networks by exploring the distribution of unlabeled data for more powerful representations. We will discuss emerging topics by revealing the intrinsic connections between unsupervised and semi-supervised learning, and propose in future directions to bridge the algorithmic and theoretical gap between transformation equivariance for unsupervised learning and supervised invariance for supervised learning, and unify unsupervised pretraining and supervised finetuning. We will also provide a broader outlook of future directions to unify transformation and instance equivariances for representation learning, connect unsupervised and semi-supervised augmentations, and explore the role of the self-supervised regularization for many learning problems.

From Shallow to Deep: Exploiting Feature-Based Classifiers for Domain Adaptation in Semantic Segmentation

The remarkable performance of Convolutional Neural Networks on image segmentation tasks comes at the cost of a large amount of pixelwise annotated images that have to be segmented for training. In contrast, feature-based learning methods, such as the Random Forest, require little training data, but rarely reach the segmentation accuracy of CNNs. This work bridges the two approaches in a transfer learning setting. We show that a CNN can be trained to correct the errors of the Random Forest in the source domain and then be applied to correct such errors in the target domain without retraining, as the domain shift between the Random Forest predictions is much smaller than between the raw data. By leveraging a few brushstrokes as annotations in the target domain, the method can deliver segmentations that are sufficiently accurate to act as pseudo-labels for target-domain CNN training. We demonstrate the performance of the method on several datasets with the challenging tasks of mitochondria, membrane and nuclear segmentation. It yields excellent performance compared to microscopy domain adaptation baselines, especially when a significant domain shift is involved.