Instance-based Deep Transfer Learning on Cross-domain Image Captioning

In recent years, large scale datasets of paired images and sentences have enabled the remarkable success in automatically generating descriptions for images, namely image captioning. However, it is labour-intensive and time-consuming to collect a sufficient number of paired images and sentences in each domain. It may be beneficial to transfer the image captioning model trained in an existing domain with pairs of images and sentences (i.e., source domain) to a new domain with only unpaired data (i.e., target domain). In this paper, we propose a cross-modal retrieval aided approach to cross-domain image captioning that leverages a cross-modal retrieval model to generate pseudo pairs of images and sentences in the target domain to facilitate the adaptation of the captioning model. To learn the correlation between images and sentences in the target domain, we propose an iterative cross-modal retrieval process where a cross-modal retrieval model is first pre-trained using the source domain data and then applied to the target domain data to acquire an initial set of pseudo image-sentence pairs. The pseudo image-sentence pairs are further refined by iteratively fine-tuning the retrieval model with the pseudo image-sentence pairs and updating the pseudo image-sentence pairs using the retrieval model. To make the linguistic patterns of the sentences learned in the source domain adapt well to the target domain, we propose an adaptive image captioning model with a self-attention mechanism fine-tuned using the refined pseudo image-sentence pairs. Experimental results on several settings where MSCOCO is used as the source domain and five different datasets (Flickr30k, TGIF, CUB-200, Oxford-102 and Conceptual) are used as the target domains demonstrate that our method achieves mostly better or comparable performance against the state-of-the-art methods. We also extend our method to cross-domain video captioning where MSR-VTT is used as the source domain and two other datasets (MSVD and Charades Captions) are used as the target domains to further demonstrate the effectiveness of our method.

In an indoor positioning system (IPS), transfer learning (TL) methods are commonly used to predict the location of mobile devices under the assumption that all training instances of the target domain are given in advance. However, this assumption has been criticized for its shortcomings in dealing with the problem of signal distribution variations, especially in a dynamic indoor environment. The reasons are: collecting a sufficient number of training instances is costly, the training instances may arrive online, the feature spaces of the target and source domains may be different, and negative knowledge may be transferred in the case of a redundant source domain. In this work, we proposed an online heterogeneous transfer learning (OHetTLAL) algorithm for IPS-based RSS fingerprinting to improve the positioning performance in the target domain by fusing both source and target domain knowledge. The source domain was refined based on the target domain to avoid negative knowledge transfer. The co-occurrence measure of the feature spaces (Cmip) was used to derive the homogeneous new feature spaces, and the features with higher weight values were selected for training the classifier because they could positively affect the location prediction of the target. Thus, the objective function was minimized over the new feature spaces. Extensive experiments were conducted on two real-world scenarios of datasets, and the predictive power of the different modeling techniques were evaluated for predicting the location of a mobile device. The results have revealed that the proposed algorithm outperforms the state-of-the-art methods for fingerprint-based indoor positioning and is found robust to changing environments. Moreover, the proposed algorithm is not only resilient to fluctuating environments but also mitigates the model's overfitting problem.

In pattern classification, we may have a few labeled data points in the target domain, but a number of labeled samples are available in another related domain (called the source domain). Transfer learning can solve such classification problems via the knowledge transfer from source to target domains. The source and target domains can be represented by heterogeneous features. There may exist uncertainty in domain transformation, and such uncertainty is not good for classification. The effective management of uncertainty is important for improving classification accuracy. So, a new belief-based bidirectional transfer classification (BDTC) method is proposed. In BDTC, the intraclass transformation matrix is estimated at first for mapping the patterns from source to target domains, and this matrix can be learned using the labeled patterns of the same class represented by heterogeneous domains (features). The labeled patterns in the source domain are transferred to the target domain by the corresponding transformation matrix. Then, we learn a classifier using all the labeled patterns in the target domain to classify the objects. In order to take full advantage of the complementary knowledge of different domains, we transfer the query patterns from target to source domains using the K-NN technique and do the classification task in the source domain. Thus, two pieces of classification results can be obtained for each query pattern in the source and target domains, but the classification results may have different reliabilities/weights. A weighted combination rule is developed to combine the two classification results based on the belief functions theory, which is an expert at dealing with uncertain information. We can efficiently reduce the uncertainty of transfer classification via the combination strategy. Experiments on some domain adaptation benchmarks show that our method can effectively improve classification accuracy compared with other related methods.

Recent progress in machine learning has been mainly due to the availability of large amounts of annotated data used for training complex models with deep architectures. Annotating this training data becomes burdensome and creates a major bottleneck in maintaining machine-learning databases. Moreover, these trained models fail to generalize to new categories or new varieties of the same categories. This is because new categories or new varieties have data distribution different from the training data distribution. To tackle these problems, this thesis proposes to develop a family of transfer-learning techniques that can deal with different training (source) and testing (target) distributions with the assumption that the availability of annotated data is limited in the testing domain. This is done by using the auxiliary data-abundant source domain from which useful knowledge is transferred that can be applied to data-scarce target domain. This transferable knowledge serves as a prior that biases target-domain predictions and prevents the target-domain model from overfitting. Specifically, we explore structural priors that encode relational knowledge between different data entities, which provides more informative bias than traditional priors. The choice of the structural prior depends on the information availability and the similarity between the two domains. Depending on the domain similarity and the information availability, we divide the transfer learning problem into four major categories and propose different structural priors to solve each of these sub-problems. This thesis first focuses on the unsupervised-domain-adaptation problem, where we propose to minimize domain discrepancy by transforming labeled source-domain data to be close to unlabeled target-domain data. For this problem, the categories remain the same across the two domains and hence we assume that the structural relationship between the source-domain samples is carried over to the target domain. Thus, graph or hyper-graph is constructed as the structural prior from both domains and a graph/hyper-graph matching formulation is used to transform samples in the source domain to be closer to samples in the target domain. An efficient optimization scheme is then proposed to tackle the time and memory inefficiencies associated with the matching problem. The few-shot learning problem is studied next, where we propose to transfer knowledge from source-domain categories containing abundantly labeled data to novel categories in the target domain that contains only few labeled data. The knowledge transfer biases the novel category predictions and prevents the model from overfitting. The knowledge is encoded using a neural-network-based prior that transforms a data sample to its corresponding class prototype. This neural network is trained from the source-domain data and applied to the target-domain data, where it transforms the few-shot samples to the novel-class prototypes for better recognition performance. The few-shot learning problem is then extended to the situation, where we do not have access to the source-domain data but only have access to the source-domain class prototypes. In this limited information setting, parametric neural-network-based priors would overfit to the source-class prototypes and hence we seek a non-parametric-based prior using manifolds. A piecewise linear manifold is used as a structural prior to fit the source-domain-class prototypes. This structure is extended to the target domain, where the novel-class prototypes are found by projecting the few-shot samples onto the manifold. Finally, the zero-shot learning problem is addressed, which is an extreme case of the few-shot learning problem where we do not have any labeled data in the target domain. However, we have high-level information for both the source and target domain categories in the form of semantic descriptors. We learn the relation between the sample space and the semantic space, using a regularized neural network so that classification of the novel categories can be carried out in a common representation space. This same neural network is then used in the target domain to relate the two spaces. In case we want to generate data for the novel categories in the target domain, we can use a constrained generative adversarial network instead of a traditional neural network. Thus, we use structural priors like graphs, neural networks and manifolds to relate various data entities like samples, prototypes and semantics for these different transfer learning sub-problems. We explore additional post-processing steps like pseudo-labeling, domain adaptation and calibration and enforce algorithmic and architectural constraints to further improve recognition performance. Experimental results on standard transfer learning image recognition datasets produced competitive results with respect to previous work. Further experimentation and analyses of these methods provided better understanding of machine learning as well.

Domain-transfer learning is a machine learning task to explore a source domain data set to help the learning problem in a target domain. Usually, the source domain has sufficient labeled data, while the target domain does not. In this paper, we propose a novel domain-transfer convolutional model by mapping a target domain data sample to a proxy in the source domain and applying a source domain model to the proxy for the purpose of prediction. In our framework, we firstly represent both source and target domains to feature vectors by two convolutional neural networks and then construct a proxy for each target domain sample in the source domain space. The proxy is supposed to be matched to the corresponding target domain sample convolutional representation vector well. To measure the matching quality, we proposed to maximize their squared-loss mutual information (SMI) between the proxy and target domain samples. We further develop a novel neural SMI estimator based on a parametric density ratio estimation function. Moreover, we also propose to minimize the classification error of both source domain samples and target domain proxies. The classification responses are also smoothened by manifolds of both the source domain and proxy space. By minimizing an objective function of SMI, classification error, and manifold regularization, we learn the convolutional networks of both source and target domains. In this way, the proxy of a target domain sample can be matched to the source domain data and thus benefits from the rich supervision information of the source domain. We design an iterative algorithm to update the parameters alternately and test it over benchmark data sets of abnormal behavior detection in video, Amazon product reviews sentiment analysis, etc.

Preoperative diagnosis of gastric cancer and primary gastric lymphoma is challenging and has important clinical significance. Inspired by the inductive reasoning learning of the human brain, transfer learning can improve diagnosis performance of target task by utilizing the knowledge learned from the other domains (source domain). However, most studies focus on single-source transfer learning and may lead to model performance degradation when a large domain shift exists between the single-source domain and target domain. By simulating the multi-modal information learning and transfer mechanism of human brain, this study designed a multisource transfer learning feature extraction and classification framework, which can enhance the prediction performance of the target model by using multisource medical data (domain). First, this manuscript designs a feature extraction network that takes the maximum mean difference based on the Wasserstein distance as an adaptive measure of probability distribution and extracts the domain-specific invariant representations between source and target domain data. Then, aiming at the random generation of parameters bringing uncertainties to prediction accuracy and generalization ability of extreme learning machine network, the 1-norm regularization is used to implement sparse constraints of the output weight matrix and improve the robustness of the model. Finally, some experiments are carried out on the data of two medical centers. The experimental results show that the area under curves (AUCs) of the method are 0.958 and 0.929 in the two validation cohorts, respectively. The method in this manuscript can provide doctors with a better diagnostic reference, which has certain practical significance.

Unsupervised domain adaptation aims to transfer knowledge from labeled source domain to unlabeled target domain. Recently, multisource domain adaptation (MDA) has begun to attract attention. Its performance should go beyond simply mixing all source domains together for knowledge transfer. In this article, we propose a novel prototype-based method for MDA. Specifically, for solving the problem that the target domain has no label, we use the prototype to transfer the semantic category information from source domains to target domain. First, a feature extraction network is applied to both source and target domains to obtain the extracted features from which the domain-invariant features and domain-specific features will be disentangled. Then, based on these two kinds of features, the named inherent class prototypes and domain prototypes are estimated, respectively. Then a prototype mapping to the extracted feature space is learned in the feature reconstruction process. Thus, the class prototypes for all source and target domains can be constructed in the extracted feature space based on the previous domain prototypes and inherent class prototypes. By forcing the extracted features are close to the corresponding class prototypes for all domains, the feature extraction network is progressively adjusted. In the end, the inherent class prototypes are used as a classifier in the target domain. Our contribution is that through the inherent class prototypes and domain prototypes, the semantic category information from source domains is transformed into the target domain by constructing the corresponding class prototypes. In our method, all source and target domains are aligned twice at the feature level for better domain-invariant features and more closer features to the class prototypes, respectively. Several experiments on public data sets also prove the effectiveness of our method.

The concept of Industry 4.0 uses cyber-physical systems and the Internet of Things to create "smart factories" that enable automated and connected production. However, the complex and diverse nature of cyber-physical systems makes them vulnerable to physical threats, where any fault in their physical components can lead to serious consequences, from financial losses to human safety. To reduce the risk of such failures, it is essential to have techniques in place to monitor the condition of these physical components closely. In recent years, data-driven predictive maintenance methods have become common due to the growth in data and advancements in machine learning. This thesis focuses on fault diagnosis, a key part of predictive maintenance, which aims to detect and identify faults in physical components. Incorporating advanced techniques, including machine learning algorithms and sensor data analysis, enhanced fault diagnosis systems' accuracy and efficiency. However, several real-world barriers challenge the effectiveness of these techniques. This research is structured around three primary objectives. The first objective aims to establish an efficient fault diagnosis system that is highly sensitive to fault occurrences while being resilient to uncertainties such as variability in fault patterns and data imbalance. To address variability in fault patterns, we developed a multi-step approach integrating clustering techniques and an adaptive neuro-fuzzy inference system, which significantly improved diagnosis robustness even under noisy conditions. For imbalanced datasets, a two-step method was introduced, combining permutation entropy for fault detection and envelope analysis for fault isolation. This approach was enhanced by an automatic frequency band selection method leveraging wavelet transform and statistical criteria, resulting in high fault diagnosis accuracy and reliability, even with limited labeled faulty data. The second objective focuses on developing fault diagnosis systems that can maintain high performance despite domain shifts and limited training data in both source and target domains. To address the challenge of domain shifts, a transfer learning technique was introduced, and a systematic literature review was conducted. This study bridges a technical gap by providing a comprehensive taxonomy, state-of-the-art methods, and insights into the challenges of applying transfer learning to predictive maintenance. Additionally, a conceptual framework was proposed to explore scenarios for transferring diagnostic knowledge from different source domains to target domains with insufficient training datasets. In a subsequent study, a new diagnostic framework was developed using digital twins and partial transfer learning to address the lack of real training datasets in both source and target domains as well as label space inconsistencies between them. The framework includes a general approach for developing digital twins of real machines to generate synthetic diagnostic data as source domain data. Furthermore, by incorporating a deep learning network and a double-layer attention mechanism, a partial transfer learning diagnostic model was developed to transfer diagnostic knowledge from the digital twin to the real target domain despite label space inconsistencies. The proposed framework was validated through an industrial case study involving rolling bearings in rotating equipment. The third objective focuses on developing an efficient fault diagnosis system capable of maintaining reliability under unseen and evolving working conditions. To achieve this, a novel self-adaptive fault diagnosis system was proposed, incorporating the MAPE-K loop, a domain generalization network model, and digital twin technologies. This self-adaptive fault diagnosis system leverages the MAPE-K loop to adapt in real time to evolving working conditions. Digital twins were employed to continuously generate data aligned with the real system’s operational context, enabling accurate fault diagnosis in dynamic environments. The domain generalization network model enhances generalization to unseen target working conditions by addressing overfitting and feature loss through multi-domain data augmentation, adversarial learning, and domain discrepancy modules. Validation using multiple rotating machinery datasets demonstrated that the developed fault diagnosis system outperformed state-of-the-art methods across various scenarios. In summary, this dissertation proposes a comprehensive and practical approach to addressing variability in fault patterns, imbalance datasets, data scarcity, domain shifts, and unseen and evolving working conditions challenges in existing fault diagnosis systems.

Multi-source adversarial transfer learning for ultrasound image segmentation with limited similarity

Many sleep studies suffer from the problem of insufficient data to fully utilize deep neural networks as different labs use different recordings set ups, leading to the need of training automated algorithms on rather small databases, whereas large annotated databases are around but cannot be directly included into these studies for data compensation due to channel mismatch. This work presents a deep transfer learning approach to overcome the channel mismatch problem and transfer knowledge from a large dataset to a small cohort to study automatic sleep staging with single-channel input. We employ the state-of-the-art SeqSleepNet and train the network in the source domain, i.e. the large dataset. Afterwards, the pretrained network is finetuned in the target domain, i.e. the small cohort, to complete knowledge transfer. We study two transfer learning scenarios with slight and heavy channel mismatch between the source and target domains. We also investigate whether, and if so, how finetuning entirely or partially the pretrained network would affect the performance of sleep staging on the target domain. Using the Montreal Archive of Sleep Studies (MASS) database consisting of 200 subjects as the source domain and the Sleep-EDF Expanded database consisting of 20 subjects as the target domain in this study, our experimental results show significant performance improvement on sleep staging achieved with the proposed deep transfer learning approach. Furthermore, these results also reveal the essential of finetuning the feature-learning parts of the pretrained network to be able to bypass the channel mismatch problem.

This paper addresses the problem of land-cover map updating by classification of multitemporal remote-sensing images in the context of domain adaptation (DA). The basic assumptions behind the proposed approach are twofold. The first one is that training data (ground reference information) are available for one of the considered multitemporal acquisitions (source domain) whereas they are not for the other (target domain). The second one is that multitemporal acquisitions (i.e., target and source domains) may be characterized by different sets of classes. Unlike other approaches available in the literature, the proposed DA Bayesian classifier based on maximum a posteriori decision rule (DA-MAP) automatically identifies whether there exist differences between the set of classes in the target and source domains and properly handles these differences in the updating process. The proposed method was tested in different scenarios of increasing complexity related to multitemporal image classification. Experimental results on medium-resolution and very high resolution multitemporal remote-sensing data sets confirm the effectiveness and the reliability of the proposed DA-MAP classifier.

As an important and widely studied research topic, the detection of fresh milk plays a crucial role in protecting consumer health. However, due to the diversity of milk brands and environmental conditions in real-world application scenarios, the accuracy of existing detection models often drops significantly when applied to new domains. A significant amount of new data needs to be gathered in order to retrain the model for new domain, which greatly increases both time and labor costs. To tackle this problem, this paper introduces a transfer learning-based cross-domain milk freshness detection method. The method uses transfer learning to generate data for unknown categories within the target domain. Specifically, the transfer generation model trains on data from both the target and source domain categories. With the proposed model, data from other categories in the source domain can be utilized to generate corresponding target domain data. This method comprehensively considers data generation from multiple perspectives of time, frequency, and spatial, to enhance the authenticity of the generated data. It uses a transfer generation architecture, consisting of a transfer network and a decoder, to learn the mapping relationship between the source and target domain, helping to narrow the gap between domains. Additionally, a feature subspace decomposition method is introduced into the decoder, and a cross-domain consistency loss function is formulated to strengthen the model’s learning capability. The proposed method is shown to generate high-quality data for unknown target-domain categories in multiple cross-domain settings, leading to a performance improvement ranging from 17.64% to 32.44% in the detection of cross-domain milk freshness.

Multisource unsupervised domain adaptation (MUDA) is an important and challenging topic for target classification with the assistance of labeled data in source domains. When we have several labeled source domains, it is difficult to map all source domains and target domain into a common feature space for classifying the targets well. In this article, a new progressive multisource domain adaptation network (PMSDAN) is proposed to further improve the classification performance. PMSDAN mainly consists of two steps for distribution alignment. First, the multiple source domains are integrated as one auxiliary domain to match the distribution with the target domain. By doing this, we can generally reduce the distribution discrepancy between each source and target domains, as well as the discrepancy between different source domains. It can efficiently explore useful knowledge from the integrated source domain. Second, to mine assistance knowledge from each source domain as much as possible, the distribution of the target domain is separately aligned with that of each source domain. A weighted fusion method is employed to combine the multiple classification results for making the final decision. In the optimization of domain adaption, weighted hybrid maximum mean discrepancy (WHMMD) is proposed, and it considers both the interclass and intraclass discrepancies. The effectiveness of the proposed PMSDAN is demonstrated in the experiments comparing with some state-of-the-art methods.

This paper introduces a new method to solve the cross-domain recognition problem. Different from the traditional domain adaption methods which rely on a global domain shift for all classes between the source and target domains, the proposed method is more flexible to capture individual class variations across domains. By adopting a natural and widely used assumption that the data samples from the same class should lay on an intrinsic low-dimensional subspace, even if they come from different domains, the proposed method circumvents the limitation of the global domain shift, and solves the cross-domain recognition by finding the joint subspaces of the source and target domains. Specifically, given labeled samples in the source domain, we construct a subspace for each of the classes. Then we construct subspaces in the target domain, called anchor subspaces, by collecting unlabeled samples that are close to each other and are highly likely to belong to the same class. The corresponding class label is then assigned by minimizing a cost function which reflects the overlap and topological structure consistency between subspaces across the source and target domains, and within the anchor subspaces, respectively. We further combine the anchor subspaces to the corresponding source subspaces to construct the joint subspaces. Subsequently, one-versus-rest support vector machine classifiers are trained using the data samples belonging to the same joint subspaces and applied to unlabeled data in the target domain. We evaluate the proposed method on two widely used datasets: 1) object recognition dataset for computer vision tasks and 2) sentiment classification dataset for natural language processing tasks. Comparison results demonstrate that the proposed method outperforms the comparison methods on both datasets.

Transductive transfer learning based Genetic Programming for balanced and unbalanced document classification using different types of features