To save content items to your account,
please confirm that you agree to abide by our usage policies.
If this is the first time you use this feature, you will be asked to authorise Cambridge Core to connect with your account.
Find out more about saving content to .
To save content items to your Kindle, first ensure firstname.lastname@example.org
is added to your Approved Personal Document E-mail List under your Personal Document Settings
on the Manage Your Content and Devices page of your Amazon account. Then enter the ‘name’ part
of your Kindle email address below.
Find out more about saving to your Kindle.
Note you can select to save to either the @free.kindle.com or @kindle.com variations.
‘@free.kindle.com’ emails are free but can only be saved to your device when it is connected to wi-fi.
‘@kindle.com’ emails can be delivered even when you are not connected to wi-fi, but note that service fees apply.
Artificial intelligence (AI) represents emerging technology with beneficial potential for the maritime domain, to contain all natural and manmade features, events, or activities appertaining to the seas, oceans or other navigable waterways. It is not a single technology but a continuum of capabilities designed to synergize computational processing power with human creativity. This chapter introduces key AI concepts, including but not limited to, algorithms, reinforcement learning, deep learning and artificial general intelligence. Science has not yet achieved sentient machines, and fully autonomous vessels may not become commonplace for a number of years; nevertheless, current AI technologies offer risk-reduction methodologies to human-crewed vessels operating in dynamic and often dangerous conditions. In general, AI can enhance compliance with the law of the sea and reduce marine casualties. Specifically, this chapter proposes that AI technologies should be adopted to facilitate safer navigation through improved hydrographic services and AI-supported decision-making for vessel masters and human crews at sea.
Exposure to ground-level ozone is a concern for both humans and vegetation, so accurate prediction of ozone time series is of great importance. However, conventional as well as emerging methods have deficiencies in predicting time series when a superposition of differently pronounced oscillations on various time scales is present. In this paper, we propose a meteorologically motivated filtering method of time series data, which can separate oscillation patterns, in combination with different multibranch neural networks. To avoid phase shifts introduced by using a causal filter, we combine past observation data with a climatological estimate about the future to be able to apply a noncausal filter in a forecast setting. In addition, the forecast in the form of the expected climatology provides some a priori information that can support the neural network to focus not merely on learning a climatological statistic. We apply this method to hourly data obtained from over 50 different monitoring stations in northern Germany situated in rural or suburban surroundings to generate a prediction for the daily maximum 8-hr average values of ground-level ozone 4 days into the future. The data preprocessing with time filters enables simpler neural networks such as fully connected networks as well as more sophisticated approaches such as convolutional and recurrent neural networks to better recognize long-term and short-term oscillation patterns like the seasonal cycle and thus leads to an improvement in the forecast skill, especially for a lead time of more than 48 hr, compared to persistence, climatological reference, and other reference models.
The selection of high-quality sperms is critical to intracytoplasmic sperm injection, which accounts for 70–80% of in vitro fertilization (IVF) treatments. So far, sperm screening is usually performed manually by clinicians. However, the performance of manual screening is limited in its objectivity, consistency, and efficiency. To overcome these limitations, we have developed a fast and noninvasive three-stage method to characterize morphology of freely swimming human sperms in bright-field microscopy images using deep learning models. Specifically, we use an object detection model to identify sperm heads, a classification model to select in-focus images, and a segmentation model to extract geometry of sperm heads and vacuoles. The models achieve an F1-score of 0.951 in sperm head detection, a z-position estimation error within ±1.5 μm in in-focus image selection, and a Dice score of 0.948 in sperm head segmentation, respectively. Customized lightweight architectures are used for the models to achieve real-time analysis of 200 frames per second. Comprehensive morphological parameters are calculated from sperm head geometry extracted by image segmentation. Overall, our method provides a reliable and efficient tool to assist clinicians in selecting high-quality sperms for successful IVF. It also demonstrates the effectiveness of deep learning in real-time analysis of live bright-field microscopy images.
There is no doubt that students’ emotional states influence their general well-being and their learning success. Although current advances of computer hardware and Artificial Intelligence (AI) techniques made it possible to include real-time emotion-sensing as one of the ways to improve students’ learning experience and performance, there are many challenges that might inhibit or delay the deployment and usage of emotional learning analytics (LA) in education. In this chapter, we will critically review the current state-of-the art of emotion detection techniques, analysis and visualisation, the benefits of emotion analysis in education and the ethical issues surrounding emotion-aware systems in education. Finally, we hope that our guidelines on how to tackle each of those issues can support research in this area.
We present a fast and precise deep-learning architecture, which we term O-Net, for obtaining super-resolved images from conventional phase-modulated optical microscopical techniques, such as phase-contrast microscopy and differential interference contrast microscopy. O-Net represents a novel deep convolutional neural network that can be trained on both simulated and experimental data, the latter of which is being demonstrated in the present context. The present study demonstrates the ability of the proposed method to achieve super-resolved images even under poor signal-to-noise ratios and does not require prior information on the point spread function or optical character of the system. Moreover, unlike previous state-of-the-art deep neural networks (such as U-Nets), the O-Net architecture seemingly demonstrates an immunity to network hallucination, a commonly cited issue caused by network overfitting when U-Nets are employed. Models derived from the proposed O-Net architecture are validated through empirical comparison with a similar sample imaged via scanning electron microscopy (SEM) and are found to generate ultra-resolved images which came close to that of the actual SEM micrograph.
Livestock welfare assessment helps monitor animal health status to maintain productivity, identify injuries and stress, and avoid deterioration. It has also become an important marketing strategy since it increases consumer pressure for a more humane transformation in animal treatment. Common visual welfare practices by professionals and veterinarians may be subjective and cost-prohibitive, requiring trained personnel. Recent advances in remote sensing, computer vision, and artificial intelligence (AI) have helped developing new and emerging technologies for livestock biometrics to extract key physiological parameters associated with animal welfare. This review discusses the livestock farming digital transformation by describing (i) biometric techniques for health and welfare assessment, (ii) livestock identification for traceability and (iii) machine and deep learning application in livestock to address complex problems. This review also includes a critical assessment of these topics and research done so far, proposing future steps for the deployment of AI models in commercial farms. Most studies focused on model development without applications or deployment for the industry. Furthermore, reported biometric methods, accuracy, and machine learning approaches presented some inconsistencies that hinder validation. Therefore, it is required to develop more efficient, non-contact and reliable methods based on AI to assess livestock health, welfare, and productivity.
Digital engineering is increasingly established in the industrial routine. Especially the application of machine learning on geometry data is a growing research issue. Driven by this, the paper presents a new method for the classification of mechanical components, which utilizes the projection of points onto a spherical detector surfaces to transfer the geometries into matrices. These matrices are then classified using deep learning networks. Different types of projection are examined, as are several deep learning models. Finally, a benchmark dataset is used to demonstrate the competitiveness.
Modern low-altitude unmanned aircraft (UA) detection and surveillance systems mostly adopt the multi-sensor fusion technology scheme of radar, visible light, infrared, acoustic and radio detection. Firstly, this paper summarises the latest research progress of UA and bird target detection and recognition technology based on radar, and provides an effective way of detection and recognition from the aspects of echo modeling and micro motion characteristic cognition, manoeuver feature enhancement and extraction, motion trajectory difference, deep learning intelligent classification, etc. Furthermore, this paper also analyses the target feature extraction and recognition algorithms represented by deep learning for other kinds of sensor data. Finally, after a comparison of the detection ability of various detection technologies, a technical scheme for low-altitude UA surveillance system based on four types of sensors is proposed, with a detailed description of its main performance indicators.
Recent years have seen a growing number of publications that analyse Natural Language Understanding (NLU) datasets for superficial cues, whether they undermine the complexity of the tasks underlying those datasets and how they impact those models that are optimised and evaluated on this data. This structured survey provides an overview of the evolving research area by categorising reported weaknesses in models and datasets and the methods proposed to reveal and alleviate those weaknesses for the English language. We summarise and discuss the findings and conclude with a set of recommendations for possible future research directions. We hope that it will be a useful resource for researchers who propose new datasets to assess the suitability and quality of their data to evaluate various phenomena of interest, as well as those who propose novel NLU approaches, to further understand the implications of their improvements with respect to their model’s acquired capabilities.
In this introductory chapter, we outline the ways in which various problems in data analysis can be formulated as optimization problems. Specifically, we discuss least squares problems, problems in matrix optimization (particularly those involving low-rank matrices), linear and kernel support vector machines, binary and multiclass logistic regression, and deep learning. We also outline the scope of the remainder of the book.
Detection of RNA spots in single-molecule fluorescence in-situ hybridization microscopy images remains a difficult task, especially when applied to large volumes of data. The variable intensity of RNA spots combined with the high noise level of the images often requires manual adjustment of the spot detection thresholds for each image. In this work, we introduce DeepSpot, a Deep Learning-based tool specifically designed for RNA spot enhancement that enables spot detection without the need to resort to image per image parameter tuning. We show how our method can enable downstream accurate spot detection. DeepSpot’s architecture is inspired by small object detection approaches. It incorporates dilated convolutions into a module specifically designed for context aggregation for small object and uses Residual Convolutions to propagate this information along the network. This enables DeepSpot to enhance all RNA spots to the same intensity, and thus circumvents the need for parameter tuning. We evaluated how easily spots can be detected in images enhanced with our method by testing DeepSpot on 20 simulated and 3 experimental datasets, and showed that accuracy of more than 97% is achieved. Moreover, comparison with alternative deep learning approaches for mRNA spot detection (deepBlink) indicated that DeepSpot provides more precise mRNA detection. In addition, we generated single-molecule fluorescence in-situ hybridization images of mouse fibroblasts in a wound healing assay to evaluate whether DeepSpot enhancement can enable seamless mRNA spot detection and thus streamline studies of localized mRNA expression in cells.
This chapter reviews recent developments in modern soft computing models, including heuristic algorithms, extreme learning machines and models based on deep learning strategies applied to water management. In this context, we describe the basics and fundamentals of the mentioned soft computing methods. We then provide a brief review of the models applied in three fields of water management: drought forecasting, evapotranspiration modelling and rainfall-runoff simulation. Thus, we provide guidelines for modern soft computing techniques applied to water management.
Drug discovery and development pipelines are timely consuming and expensive, depending on numerous factors. Artificial intelligence (AI) tools are increasingly being applied in drug discovery for Alzheimer’s disease (AD). In the “big data” era, AI offers cutting-edge applications of informatics and computational tools for re-defining disease biology, discovering new therapeutics, and identifying novel targets with the least errors. The application of AI has the potential to enhance the pipeline across all stages of drug discovery and reduce failure rates in drug development for AD. In this chapter, we introduce AI techniques accessible for accelerating drug discovery. We summarize representation learning, machine learning, and deep learning toolboxes, available for drug discovery. We illustrate the application of AI for target identification, evaluation of pharmacokinetic properties (i.e., brain penetration), safety, and identification of biomarkers in clinical trials. We discuss current challenges and future directions of AI-based solutions for drug discovery. Rapidly developing, powerful and innovative AI technologies can expedite drug discovery and development for AD.
Vaginitis is a prevalent gynecologic disease that threatens millions of women’s health. Although microscopic examination of vaginal discharge is an effective method to identify vaginal infections, manual analysis of microscopic leucorrhea images is extremely time-consuming and labor-intensive. To automate the detection and identification of visible components in microscopic leucorrhea images for early-stage diagnosis of vaginitis, we propose a novel end-to-end deep learning-based cells detection framework using attention-based detection with transformers (DETR) architecture. The transfer learning was applied to speed up the network convergence while maintaining the lowest annotation cost. To address the issue of detection performance degradation caused by class imbalance, the weighted sampler with on-the-fly data augmentation module was integrated into the detection pipeline. Additionally, the multi-head attention mechanism and the bipartite matching loss system of the DETR model perform well in identifying partially overlapping cells in real-time. With our proposed method, the pipeline achieved a mean average precision (mAP) of 86.00% and the average precision (AP) of epithelium, leukocyte, pyocyte, mildew, and erythrocyte was 96.76, 83.50, 74.20, 89.66, and 88.80%, respectively. The average test time for a microscopic leucorrhea image is approximately 72.3 ms. Currently, this cell detection method represents state-of-the-art performance.
Deep learning approaches are superior in natural language processing due to their ability to extract informative features and patterns from languages. The two most successful neural architectures are LSTM and transformers, used in large pretrained language models such as BERT. While cross-lingual approaches are on the rise, most current natural language processing techniques are designed and applied to English, and less-resourced languages are lagging behind. In morphologically rich languages, information is conveyed through morphology, for example, through affixes modifying stems of words. The existing neural approaches do not explicitly use the information on word morphology. We analyse the effect of adding morphological features to LSTM and BERT models. As a testbed, we use three tasks available in many less-resourced languages: named entity recognition (NER), dependency parsing (DP) and comment filtering (CF). We construct baselines involving LSTM and BERT models, which we adjust by adding additional input in the form of part of speech (POS) tags and universal features. We compare the models across several languages from different language families. Our results suggest that adding morphological features has mixed effects depending on the quality of features and the task. The features improve the performance of LSTM-based models on the NER and DP tasks, while they do not benefit the performance on the CF task. For BERT-based models, the added morphological features only improve the performance on DP when they are of high quality (i.e., manually checked) while not showing any practical improvement when they are predicted. Even for high-quality features, the improvements are less pronounced in language-specific BERT variants compared to massively multilingual BERT models. As in NER and CF datasets manually checked features are not available, we only experiment with predicted features and find that they do not cause any practical improvement in performance.
Previous analyses of grey and white matter volumes have reported that schizophrenia is associated with structural changes. Deep learning is a data-driven approach that can capture highly compact hierarchical non-linear relationships among high-dimensional features, and therefore can facilitate the development of clinical tools for making a more accurate and earlier diagnosis of schizophrenia.
To identify consistent grey matter abnormalities in patients with schizophrenia, 662 people with schizophrenia and 613 healthy controls were recruited from eight centres across China, and the data from these independent sites were used to validate deep-learning classifiers.
We used a prospective image-based meta-analysis of whole-brain voxel-based morphometry. We also automatically differentiated patients with schizophrenia from healthy controls using combined grey matter, white matter and cerebrospinal fluid volumetric features, incorporated a deep neural network approach on an individual basis, and tested the generalisability of the classification models using independent validation sites.
We found that statistically reliable schizophrenia-related grey matter abnormalities primarily occurred in regions that included the superior temporal gyrus extending to the temporal pole, insular cortex, orbital and middle frontal cortices, middle cingulum and thalamus. Evaluated using leave-one-site-out cross-validation, the performance of the classification of schizophrenia achieved by our findings from eight independent research sites were: accuracy, 77.19–85.74%; sensitivity, 75.31–89.29% and area under the receiver operating characteristic curve, 0.797–0.909.
These results suggest that, by using deep-learning techniques, multidimensional neuroanatomical changes in schizophrenia are capable of robustly discriminating patients with schizophrenia from healthy controls, findings which could facilitate clinical diagnosis and treatment in schizophrenia.
In order to merge the advantages of the traditional compressed sensing (CS) methodology and the data-driven deep network scheme, this paper proposes a physical model-driven deep network, termed CS-Net, for solving target image reconstruction problems in through-the-wall radar imaging. The proposed method consists of two consequent steps. First, a learned convolutional neural network prior is introduced to replace the regularization term in the traditional iterative CS-based method to capture the redundancy of the radar echo signal. Moreover, the physical model of the radar signal is used in the data consistency layer to encourage consistency with the measurements. Second, the iterative CS optimization is unrolled to yield a deep learning network, where the weight, regularization parameter, and the other parameters are learnable. A quantity of training data enables the network to extract high-dimensional characteristics of the radar echo signal to reconstruct the spatial target image. Simulation results demonstrated that the proposed method can achieve accurate target image reconstruction and was superior to the traditional CS method, in terms of mean squared error and the target texture details.
Machine Reading Comprehension (MRC) is a challenging task and hot topic in Natural Language Processing. The goal of this field is to develop systems for answering the questions regarding a given context. In this paper, we present a comprehensive survey on diverse aspects of MRC systems, including their approaches, structures, input/outputs, and research novelties. We illustrate the recent trends in this field based on a review of 241 papers published during 2016–2020. Our investigation demonstrated that the focus of research has changed in recent years from answer extraction to answer generation, from single- to multi-document reading comprehension, and from learning from scratch to using pre-trained word vectors. Moreover, we discuss the popular datasets and the evaluation metrics in this field. The paper ends with an investigation of the most-cited papers and their contributions.
Reproducibility of a deep-learning fully convolutional neural network is evaluated by training several times the same network on identical conditions (database, hyperparameters, and hardware) with nondeterministic graphics processing unit operations. The network is trained to model three typical time–space-evolving physical systems in two dimensions: heat, Burgers’, and wave equations. The behavior of the networks is evaluated on both recursive and nonrecursive tasks. Significant changes in models’ properties (weights and feature fields) are observed. When tested on various benchmarks, these models systematically return estimations with a high level of deviation, especially for the recurrent analysis which strongly amplifies variability due to the nondeterminism. Trainings performed with double floating-point precision provide slightly better estimations and a significant reduction of the variability of both the network parameters and its testing error range.
Machine-generated artworks are now part of the contemporary art scene: they are attracting significant investments and they are presented in exhibitions together with those created by human artists. These artworks are mainly based on generative deep learning (GDL) techniques, which have seen a formidable development and remarkable refinement in the very recent years. Given the inherent characteristics of these techniques, a series of novel legal problems arise. In this article, we consider a set of key questions in the area of GDL for the arts, including the following: is it possible to use copyrighted works as training set for generative models? How do we legally store their copies in order to perform the training process? Who (if someone) will own the copyright on the generated data? We try to answer these questions considering the law in force in both the United States and the European Union, and potential future alternatives. We then extend our analysis to code generation, which is an emerging area of GDL. Finally, we also formulate a set of practical guidelines for artists and developers working on deep learning generated art, as well as some policy suggestions for policymakers.