UCL-Dehaze: Toward Real-World Image Dehazing via Unsupervised Contrastive Learning.

While the wisdom of training an image dehazing model on synthetic hazy data can alleviate the difficulty of collecting real-world hazy/clean image pairs, it brings the well-known domain shift problem. From a different yet new perspective, this paper explores contrastive learning with an adversarial training effort to leverage unpaired real-world hazy and clean images, thus alleviating the domain shift problem and enhancing the network's generalization ability in real-world scenarios. We propose an effective unsupervised contrastive learning paradigm for image dehazing, dubbed UCL-Dehaze. Unpaired real-world clean and hazy images are easily captured, and will serve as the important positive and negative samples respectively when training our UCL-Dehaze network. To train the network more effectively, we formulate a new self-contrastive perceptual loss function, which encourages the restored images to approach the positive samples and keep away from the negative samples in the embedding space. Besides the overall network architecture of UCL-Dehaze, adversarial training is utilized to align the distributions between the positive samples and the dehazed images. Compared with recent image dehazing works, UCL-Dehaze does not require paired data during training and utilizes unpaired positive/negative data to better enhance the dehazing performance. We conduct comprehensive experiments to evaluate our UCL-Dehaze and demonstrate its superiority over the state-of-the-arts, even only 1,800 unpaired real-world images are used to train our network. Source code is publicly available at https://github.com/yz-wang/UCL-Dehaze.

Technology-enhanced higher education: Text mining and bibliometrics.

Objectives: Research on technology-enhanced higher education (TEHE) has been active and influential in educational technology. The study had three objectives: (1) to recognize the tendencies in the field and the contributing countries/regions/institutions, (2) to visualize scientific collaborations, and (3) to reveal important research topics, their developmental tendencies, correlations, and distributions across contributing countries/regions/institutions. Methods: We collected 609 papers in relation to TEHE from 2004 to 2022 and analyzed them using text mining and bibliometric methods. Specifically, we focused on determining article trends, identifying contributing institutions/countries/regions, visualizing scientific collaborations through social network analysis, and revealing the important topics and their conceptual evolutions over time using topic models, Mann-Kendall trend test, hierarchical clustering, and Sankey visualization. Results: Regarding the first objective, TEHE articles have grown consistently and will continue to expand. This growth was due to the contributions of Spanish universities and institutions from other countries/regions such as the USA, the UK, Australia, Germany, China, and Turkey. Regarding the second objective, the exploration of regional and institutional collaborations through social networks revealed that geographically adjacent institutions tended to foster close collaborations, particularly among those sharing similar research interests. Nevertheless, more cross-regional collaborations are needed to advance TEHE research. Regarding the third objective, the analysis of topics highlighted research hotspots and emerging themes such as Massive Online Open Courses, AI and big data in education, Gamification and engagement, Learning effectiveness and strategies, Social networks and discussion forums, COVID-19 and online learning, and Plagiarism detection and learning analytics. Conclusions: This bibliometric study comprehensively analyzed the research landscape of TEHE research regarding contributors, collaborations, and research topics, and offers a glimpse into what the future may hold. It can be used as a guide for contributors to the field to identify the current research hotspots and emerging themes.

Norest-Net: Normal Estimation Neural Network for 3-D Noisy Point Clouds.

The widely deployed ways to capture a set of unorganized points, e.g., merged laser scans, fusion of depth images, and structure-from- x , usually yield a 3-D noisy point cloud. Accurate normal estimation for the noisy point cloud makes a crucial contribution to the success of various applications. However, the existing normal estimation wisdoms strive to meet a conflicting goal of simultaneously performing normal filtering and preserving surface features, which inevitably leads to inaccurate estimation results. We propose a normal estimation neural network (Norest-Net), which regards normal filtering and feature preservation as two separate tasks, so that each one is specialized rather than traded off. For full noise removal, we present a normal filtering network (NF-Net) branch by learning from the noisy height map descriptor (HMD) of each point to the ground-truth (GT) point normal; for surface feature recovery, we construct a normal refinement network (NR-Net) branch by learning from the bilaterally defiltered point normal descriptor (B-DPND) to the GT point normal. Moreover, NR-Net is detachable to be incorporated into the existing normal estimation methods to boost their performances. Norest-Net shows clear improvements over the state of the arts in both feature preservation and noise robustness on synthetic and real-world captured point clouds.

PointSee: Image Enhances Point Cloud.

There is a prevailing trend towards fusing multi-modal information for 3D object detection (3OD). However, challenges related to computational efficiency, plug-and-play capabilities, and accurate feature alignment have not been adequately addressed in the design of multi-modal fusion networks. In this paper, we present PointSee, a lightweight, flexible, and effective multi-modal fusion solution to facilitate various 3OD networks by semantic feature enhancement of point clouds (e.g., LiDAR or RGB-D data) assembled with scene images. Beyond the existing wisdom of 3OD, PointSee consists of a hidden module (HM) and a seen module (SM): HM decorates point clouds using 2D image information in an offline fusion manner, leading to minimal or even no adaptations of existing 3OD networks; SM further enriches the point clouds by acquiring point-wise representative semantic features, leading to enhanced performance of existing 3OD networks. Besides the new architecture of PointSee, we propose a simple yet efficient training strategy, to ease the potential inaccurate regressions of 2D object detection networks. Extensive experiments on the popular outdoor/indoor benchmarks show quantitative and qualitative improvements of our PointSee over thirty-five state-of-the-art methods.

Asymmetric cross-modal attention network with multimodal augmented mixup for medical visual question answering.

Insufficient training data is a common barrier to effectively learn multimodal information interactions and question semantics in existing medical Visual Question Answering (VQA) models. This paper proposes a new Asymmetric Cross Modal Attention network called ACMA, which constructs an image-guided attention and a question-guided attention to improve multimodal interactions from insufficient data. In addition, a Semantic Understanding Auxiliary (SUA) in the question-guided attention is newly designed to learn rich semantic embeddings for improving model performance on question understanding by integrating word-level and sentence-level information. Moreover, we propose a new data augmentation method called Multimodal Augmented Mixup (MAM) to train the ACMA, denoted as ACMA-MAM. The MAM incorporates various data augmentations and a vanilla mixup strategy to generate more non-repetitive data, which avoids time-consuming artificial data annotations and improves model generalization capability. Our ACMA-MAM outperforms state-of-the-art models on three publicly accessible medical VQA datasets (VQA-Rad, VQA-Slake, and PathVQA) with accuracies of 76.14 %, 83.13 %, and 53.83 % respectively, achieving improvements of 2.00 %, 1.32 %, and 1.59 % accordingly. Moreover, our model achieves F1 scores of 78.33 %, 82.83 %, and 51.86 %, surpassing the state-of-the-art models by 2.80 %, 1.15 %, and 1.37 % respectively.

Contrastive learning-guided multi-meta attention network for breast ultrasound video diagnosis.

Breast cancer is the most common cause of cancer death in women. Early screening and treatment can effectively improve the success rate of treatment. Ultrasound imaging technology, as the preferred modality for breast cancer screening, provides an essential reference for early diagnosis. Existing computer-aided ultrasound imaging diagnostic techniques mainly rely on the selected key frames for breast cancer lesion diagnosis. In this paper, we first collected and annotated a dataset of ultrasound video sequences of 268 cases of breast lesions. Moreover, we propose a contrastive learning-guided multi-meta attention network (CLMAN) by combining a deformed feature extraction module and a multi-meta attention module to address breast lesion diagnosis in ultrasound sequence. The proposed feature extraction module can autonomously acquire key information of the feature map in the spatial dimension, whereas the designed multi-meta attention module is dedicated to effective information aggregation in the temporal dimension. In addition, we utilize a contrast learning strategy to alleviate the problem of high imaging variability within ultrasound lesion videos. The experimental results on our collected dataset show that our CLMAN significantly outperforms existing advanced methods for video classification.

Machine and cognitive intelligence for human health: systematic review.

Brain informatics is a novel interdisciplinary area that focuses on scientifically studying the mechanisms of human brain information processing by integrating experimental cognitive neuroscience with advanced Web intelligence-centered information technologies. Web intelligence, which aims to understand the computational, cognitive, physical, and social foundations of the future Web, has attracted increasing attention to facilitate the study of brain informatics to promote human health. A large number of articles created in the recent few years are proof of the investment in Web intelligence-assisted human health. This study systematically reviews academic studies regarding article trends, top journals, subjects, countries/regions, and institutions, study design, artificial intelligence technologies, clinical tasks, and performance evaluation. Results indicate that literature is especially welcomed in subjects such as medical informatics and health care sciences and service. There are several promising topics, for example, random forests, support vector machines, and conventional neural networks for disease detection and diagnosis, semantic Web, ontology mining, and topic modeling for clinical or biomedical text mining, artificial neural networks and logistic regression for prediction, and convolutional neural networks and support vector machines for monitoring and classification. Additionally, future research should focus on algorithm innovations, additional information use, functionality improvement, model and system generalization, scalability, evaluation, and automation, data acquirement and quality improvement, and allowing interaction. The findings of this study help better understand what and how Web intelligence can be applied to promote healthcare procedures and clinical outcomes. This provides important insights into the effective use of Web intelligence to support informatics-enabled brain studies.

TACN: A Topical Adversarial Capsule Network for textual network embedding.

Combining topological information and attributed information of nodes in networks effectively is a valuable task in network embedding. Nevertheless, many prior network embedding methods regarded attributed information of nodes as simple attribute sets or ignored them totally. In some scenarios, the hidden information contained in vertex attributes are essential to network embedding. For instance, networks that contain vertexes with text information play an increasingly important role in our life, including citation networks, social networks, and entry networks. In these textual networks, the latent topic relevance information of different vertexes contained in textual attributes information are valuable in the network analysis process. Shared latent topics of nodes in networks may influence the interaction between them, which is critical to network embedding. However, much prior work for textual network embedding only regarded the text information as simple word sets while ignored the embedded topic information. In this paper, we develop a model named Topical Adversarial Capsule Network (TACN) for textual network embedding, which extracts a low-dimensional latent space of the original network from node structures, vertex attributes, and topic information contained in text of nodes. The proposed TACN contains three parts. The first part is an embedding model, which extracts the embedding representation from the topological structure, vertex attributes, and document-topic distributions. To ensure a consistent training process by back-propagation, we generate document-topic distributions by the neural topic model with Gaussian Softmax constructions. The second part is a prediction model, which is used to exploit labels of vertices. In the third part, an adversarial capsule model is used to help distinguish the latent representations from node structure domain, vertex attribute domain, or document-topic distribution domain. The latent representations, which may come from the three domains, are the output of the embedding model. We incorporate the adversarial idea into the adversarial capsule model to combine the information from these three domains, rather than to distinguish the representations conventionally. Experiments on seven real-world datasets validate the effectiveness of our method.

BGM-Net: Boundary-Guided Multiscale Network for Breast Lesion Segmentation in Ultrasound.

Automatic and accurate segmentation of breast lesion regions from ultrasonography is an essential step for ultrasound-guided diagnosis and treatment. However, developing a desirable segmentation method is very difficult due to strong imaging artifacts e.g., speckle noise, low contrast and intensity inhomogeneity, in breast ultrasound images. To solve this problem, this paper proposes a novel boundary-guided multiscale network (BGM-Net) to boost the performance of breast lesion segmentation from ultrasound images based on the feature pyramid network (FPN). First, we develop a boundary-guided feature enhancement (BGFE) module to enhance the feature map for each FPN layer by learning a boundary map of breast lesion regions. The BGFE module improves the boundary detection capability of the FPN framework so that weak boundaries in ambiguous regions can be correctly identified. Second, we design a multiscale scheme to leverage the information from different image scales in order to tackle ultrasound artifacts. Specifically, we downsample each testing image into a coarse counterpart, and both the testing image and its coarse counterpart are input into BGM-Net to predict a fine and a coarse segmentation maps, respectively. The segmentation result is then produced by fusing the fine and the coarse segmentation maps so that breast lesion regions are accurately segmented from ultrasound images and false detections are effectively removed attributing to boundary feature enhancement and multiscale image information. We validate the performance of the proposed approach on two challenging breast ultrasound datasets, and experimental results demonstrate that our approach outperforms state-of-the-art methods.

Target languages, types of activities, engagement, and effectiveness of extramural language learning.

Since Sundqvist introduced the term "extramural English" in 2009, empirical research on extramural language learning has continued to expand. However, the expanding empirical research has yet yielded incommensurate review studies. To present a timely picture of the field of extramural language learning, this study conducts a review of 33 relevant articles retrieved from Scopus and Web of Science databases. The results showed the five types of target languages frequently investigated in this field (i.e., English, German, French, Chinese, and Japanese) and seven main types of extramural learning activities (i.e., playing digital games, watching videos, reading, listening to audios, having technology-enhanced socialisation, having face-to-face socialisation, and writing compositions). People's engagement in extramural language learning was overall high, especially listening to audios and playing digital games, mediated by the relationship between the difficulty of the activities and people's target language proficiency levels, gender, and the interactive environment. Extramural language learning was overall effective for language development and enhancing affective states in language learning. The effectiveness may be influenced by the involvement of language inputs and outputs and the amount of engagement time. Implications for practitioners were suggested concerning encouraging digital gameplay, emphasising formal language instruction, and creating positive interactive environments for extramural language learning.

Fast Supervised Topic Models for Short Text Emotion Detection.

With the development of social network platforms, discussion forums, and question answering websites, a huge number of short messages that typically contain a few words for an individual document are posted by online users. In these short messages, emotions are frequently embedded for communicating opinions, expressing friendship, and promoting influence. It is quite valuable to detect emotions from short messages, but the corresponding task suffers from the sparsity of feature space. In this article, we first generate term groups co-occurring in the same context to enrich the number of features. Then, two basic supervised topic models are proposed to associate emotions with topics accurately. To reduce the time cost of parameter estimation, we further propose an accelerated algorithm for our basic models. Extensive evaluations using three short corpora validate the efficiency and effectiveness of the accelerated models for predicting the emotions of unlabeled documents, in addition to generate the topic-level emotion lexicons.

Temporal event searches based on event maps and relationships.

To satisfy a user's need to find and understand the whole picture of an event effectively and efficiently, in this paper we formalize the problem of temporal event searches and propose a framework of event relationship analysis for search events based on user queries. We define three kinds of event relationships: temporal, content dependence, and event reference, that can be used to identify to what extent a component event is dependent on another in the evolution of a target event (i.e., the query event). The search results are organized as a temporal event map (TEM) that serves as the whole picture about an event's evolution or development by showing the dependence relationships among events. Based on the event relationships in the TEM, we further propose a method to measure the degrees of importance of events, so as to discover the important component events for a query, as well as the several algebraic operators involved in the TEM, that allow users to view the target event. Experiments conducted on a real data set show that our method outperforms the baseline method Event Evolution Graph (EEG), and it can help discover certain new relationships missed by previous methods and even by human annotators.

Joint Learning of Multiple Differential Networks With Latent Variables.

Graphical models have been widely used to learn the conditional dependence structures among random variables. In many controlled experiments, such as the studies of disease or drug effectiveness, learning the structural changes of graphical models under two different conditions is of great importance. However, most existing graphical models are developed for estimating a single graph and based on a tacit assumption that there is no missing relevant variables, which wastes the common information provided by multiple heterogeneous data sets and underestimates the influence of latent/unobserved relevant variables. In this paper, we propose a joint differential network analysis (JDNA) model to jointly estimate multiple differential networks with latent variables from multiple data sets. The JDNA model is built on a penalized D-trace loss function, with group lasso or generalized fused lasso penalties. We implement a proximal gradient-based alternating direction method of multipliers to tackle the corresponding convex optimization problems. Extensive simulation experiments demonstrate that JDNA model outperforms state-of-the-art methods in estimating the structural changes of graphical models. Moreover, a series of experiments on several real-world data sets have been performed and experiment results consistently show that our proposed JDNA model is effective in identifying differential networks under different conditions.

A bibliometric analysis of natural language processing in medical research.

BACKGROUND: Natural language processing (NLP) has become an increasingly significant role in advancing medicine. Rich research achievements of NLP methods and applications for medical information processing are available. It is of great significance to conduct a deep analysis to understand the recent development of NLP-empowered medical research field. However, limited study examining the research status of this field could be found. Therefore, this study aims to quantitatively assess the academic output of NLP in medical research field. METHODS: We conducted a bibliometric analysis on NLP-empowered medical research publications retrieved from PubMed in the period 2007-2016. The analysis focused on three aspects. Firstly, the literature distribution characteristics were obtained with a statistics analysis method. Secondly, a network analysis method was used to reveal scientific collaboration relations. Finally, thematic discovery and evolution was reflected using an affinity propagation clustering method. RESULTS: There were 1405 NLP-empowered medical research publications published during the 10 years with an average annual growth rate of 18.39%. 10 most productive publication sources together contributed more than 50% of the total publications. The USA had the highest number of publications. A moderately significant correlation between country's publications and GDP per capita was revealed. Denny, Joshua C was the most productive author. Mayo Clinic was the most productive affiliation. The annual co-affiliation and co-country rates reached 64.04% and 15.79% in 2016, respectively. 10 main great thematic areas were identified including Computational biology, Terminology mining, Information extraction, Text classification, Social medium as data source, Information retrieval, etc. CONCLUSIONS: A bibliometric analysis of NLP-empowered medical research publications for uncovering the recent research status is presented. The results can assist relevant researchers, especially newcomers in understanding the research development systematically, seeking scientific cooperation partners, optimizing research topic choices and monitoring new scientific or technological activities.