Because learning sometimes involves sensitive data, machine learning algorithms have been extended to offer privacy for training data. In practice, this has been mostly an afterthought, with privacy-preserving models obtained by re-running training with a different optimizer, but using the model architectures that already performed well in a non-privacy-preserving setting. This approach leads to less than ideal privacy/utility tradeoffs, as we show here. Instead, we propose that model architectures are chosen ab initio explicitly for privacy-preserving training. To provide guarantees under the gold standard of differential privacy, one must bound as strictly as possible how individual training points can possibly affect model updates. In this paper, we are the first to observe that the choice of activation function is central to bounding the sensitivity of privacy-preserving deep learning. We demonstrate analytically and experimentally how a general family of bounded activation functions, the tempered sigmoids, consistently outperform unbounded activation functions like ReLU. Using this paradigm, we achieve new state-of-the-art accuracy on MNIST, FashionMNIST, and CIFAR10 without any modification of the learning procedure fundamentals or differential privacy analysis.
In recent years, we have witnessed the rise of deep learning. Deep neural networks have proved their success in many areas. However, the optimization of these networks has become more difficult as neural networks going deeper and datasets becoming bigger. Therefore, more advanced optimization algorithms have been proposed over the past years. In this study, widely used optimization algorithms for deep learning are examined in detail. To this end, these algorithms called adaptive gradient methods are implemented for both supervised and unsupervised tasks. The behaviour of the algorithms during training and results on four image datasets, namely, MNIST, CIFAR-10, Kaggle Flowers and Labeled Faces in the Wild are compared by pointing out their differences against basic optimization algorithms.
We propose an unsupervised image fusion architecture for multiple application scenarios based on the combination of multi-scale discrete wavelet transform through regional energy and deep learning. To our best knowledge, this is the first time the conventional image fusion method has been combined with deep learning. The useful information of feature maps can be utilized adequately through multi-scale discrete wavelet transform in our proposed method.Compared with other state-of-the-art fusion method, the proposed algorithm exhibits better fusion performance in both subjective and objective evaluation. Moreover, it's worth mentioning that comparable fusion performance trained in COCO dataset can be obtained by training with a much smaller dataset with only hundreds of images chosen randomly from COCO. Hence, the training time is shortened substantially, leading to the improvement of the model's performance both in practicality and training efficiency.
Emotion analysis has been attracting researchers' attention. Most previous works in the artificial intelligence field focus on recognizing emotion rather than mining the reason why emotions are not or wrongly recognized. Correlation among emotions contributes to the failure of emotion recognition. In this paper, we try to fill the gap between emotion recognition and emotion correlation mining through natural language text from web news. Correlation among emotions, expressed as the confusion and evolution of emotion, is primarily caused by human emotion cognitive bias. To mine emotion correlation from emotion recognition through text, three kinds of features and two deep neural network models are presented. The emotion confusion law is extracted through orthogonal basis. The emotion evolution law is evaluated from three perspectives, one-step shift, limited-step shifts, and shortest path transfer. The method is validated using three datasets-the titles, the bodies, and the comments of news articles, covering both objective and subjective texts in varying lengths (long and short). The experimental results show that, in subjective comments, emotions are easily mistaken as anger. Comments tend to arouse emotion circulations of love-anger and sadness-anger. In objective news, it is easy to recognize text emotion as love and cause fear-joy circulation. That means, journalists may try to attract attention using fear and joy words but arouse the emotion love instead; After news release, netizens generate emotional comments to express their intense emotions, i.e., anger, sadness, and love. These findings could provide insights for applications regarding affective interaction such as network public sentiment, social media communication, and human-computer interaction.
Considerable morphological phenotyping studies in nephrology have emerged in the past few years, aiming to discover hidden regularities between clinical and imaging phenotypes. Such studies have been largely enabled by deep learning based image analysis to extract sparsely located targeting objects (e.g., glomeruli) on high-resolution whole slide images (WSI). However, such methods need to be trained using labor-intensive high-quality annotations, ideally labeled by pathologists. Inspired by the recent "human-in-the-loop" strategy, we developed EasierPath, an open-source tool to integrate human physicians and deep learning algorithms for efficient large-scale pathological image quantification as a loop. Using EasierPath, physicians are able to (1) optimize the recall and precision of deep learning object detection outcomes adaptively, (2) seamlessly support deep learning outcomes refining using either our EasierPath or prevalent ImageScope software without changing physician's user habit, and (3) manage and phenotype each object with user-defined classes. As a user case of EasierPath, we present the procedure of curating large-scale glomeruli in an efficient human-in-the-loop fashion (with two loops). From the experiments, the EasierPath saved 57 % of the annotation efforts to curate 8,833 glomeruli during the second loop. Meanwhile, the average precision of glomerular detection was leveraged from 0.504 to 0.620. The EasierPath software has been released as open-source to enable the large-scale glomerular prototyping. The code can be found in https://github.com/yuankaihuo/EasierPath
The Gauss-Manin connection of a family of hypersurfaces governs the change of the period matrix along the family. This connection can be complicated even when the equations defining the family look simple. When this is the case, it is computationally expensive to compute the period matrices of varieties in the family via homotopy continuation. We train neural networks that can quickly and reliably guess the complexity of the Gauss-Manin connection of a pencil of hypersurfaces. As an application, we compute the periods of 96% of smooth quartic surfaces in projective 3-space whose defining equation is a sum of five monomials; from the periods of these quartic surfaces, we extract their Picard numbers and the endomorphism fields of their transcendental lattices.
Optimising deep learning inference across edge devices and optimisation targets such as inference time, memory footprint and power consumption is a key challenge due to the ubiquity of neural networks. Today, production deep learning frameworks provide useful abstractions to aid machine learning engineers and systems researchers. However, in exchange they can suffer from compatibility challenges (especially on constrained platforms), inaccessible code complexity, or design choices that otherwise limit research from a systems perspective. This paper presents Orpheus, a new deep learning framework for easy prototyping, deployment and evaluation of inference optimisations. Orpheus features a small codebase, minimal dependencies, and a simple process for integrating other third party systems. We present some preliminary evaluation results.
Wrist accelerometers for assessing hallmark measures of physical activity (PA) are rapidly growing with the advent of smartwatch technology. Given the growing popularity of wrist-worn accelerometers, there needs to be a rigorous evaluation for recognizing (PA) type and estimating energy expenditure (EE) across the lifespan. Participants (66% women, aged 20-89 yrs) performed a battery of 33 daily activities in a standardized laboratory setting while a tri-axial accelerometer collected data from the right wrist. A portable metabolic unit was worn to measure metabolic intensity. We built deep learning networks to extract spatial and temporal representations from the time-series data, and used them to recognize PA type and estimate EE. The deep learning models resulted in high performance; the F1 score was: 0.82, 0.81, and 95 for recognizing sedentary, locomotor, and lifestyle activities, respectively. The root mean square error was 1.1 (+/-0.13) for the estimation of EE.
In recent years, deep learning has made great progress in many fields such as image recognition, natural language processing, speech recognition and video super-resolution. In this survey, we comprehensively investigate 33 state-of-the-art video super-resolution (VSR) methods based on deep learning. It is well known that the leverage of information within video frames is important for video super-resolution. Thus we propose a taxonomy and classify the methods into six sub-categories according to the ways of utilizing inter-frame information. Moreover, the architectures and implementation details of all the methods are depicted in detail. Finally, we summarize and compare the performance of the representative VSR method on some benchmark datasets. We also discuss some challenges, which need to be further addressed by researchers in the community of VSR. To the best of our knowledge, this work is the first systematic review on VSR tasks, and it is expected to make a contribution to the development of recent studies in this area and potentially deepen our understanding to the VSR techniques based on deep learning.
Pervasive computing applications deal with intelligence surrounding users that can facilitate their activities. This intelligence is provided in the form of software components incorporated in embedded systems or devices in close distance with end users.One example infrastructure that can host intelligent pervasive services is the Edge Computing (EC) infrastructure. EC nodes can execute a number of tasks for data collected by devices present in the Internet of Things (IoT) infrastructure. In this paper, we propose an intelligent, proactive tasks management model based on the demand. Demand depicts the number of users or applications interested in using the available tasks in EC nodes, thus, characterizing their popularity. We rely on a Deep Machine Learning (DML) model and more specifically on a Long Short Term Memory (LSTM) network to learn the distribution of demand indicators for each task and estimate the future interest. This information is combined with historical observations and support a decision making scheme to conclude which tasks will be offloaded due to limited interest on them. We have to notice that in our decision making, we also take into consideration the load that every task may add to the processing node where it will be allocated. The description of our model is accompanied by a large set of experimental simulations for evaluating the proposed mechanism. We provide numerical results and reveal that the proposed scheme is capable of deciding on the fly while concluding the most efficient allocation.
Deep learning provides the opportunity to improve upon conflicting reports considering the relationship between the Amazon river's fish and dolphin abundance and reduced canopy cover as a result of deforestation. Current methods of fish and dolphin abundance estimates are performed by on-site sampling using visual and capture/release strategies. We propose a novel approach to calculating fish abundance using deep learning for fish and dolphin estimates from sonar images taken from the back of a trolling boat. We consider a data set of 143 images ranging from 0-34 fish, and 0-3 dolphins provided by the Fund Amazonia research group. To overcome the data limitation, we test the capabilities of data augmentation on an unconventional 15/85 training/testing split. Using 20 training images, we simulate a gradient of data up to 25,000 images using augmented backgrounds and randomly placed/rotation cropped fish and dolphin taken from the training set. We then train four multitask network architectures: DenseNet201, InceptionNetV2, Xception, and MobileNetV2 to predict fish and dolphin numbers using two function approximation methods: regression and classification. For regression, Densenet201 performed best for fish and Xception best for dolphin with mean squared errors of 2.11 and 0.133 respectively. For classification, InceptionResNetV2 performed best for fish and MobileNetV2 best for dolphins with a mean error of 2.07 and 0.245 respectively. Considering the 123 testing images, our results show the success of data simulation for limited sonar data sets. We find DenseNet201 is able to identify dolphins after approximately 5000 training images, while fish required the full 25,000. Our method can be used to lower costs and expedite the data analysis of fish and dolphin abundance to real-time along the Amazon river and river systems worldwide.
Deep Learning-based object detectors can enhance the capabilities of smart camera systems in a wide spectrum of machine vision applications including video surveillance, autonomous driving, robots and drones, smart factory, and health monitoring. Pedestrian detection plays a key role in all these applications and deep learning can be used to construct accurate state-of-the-art detectors. However, such complex paradigms do not scale easily and are not traditionally implemented in resource-constrained smart cameras for on-device processing which offers significant advantages in situations when real-time monitoring and robustness are vital. Efficient neural networks can not only enable mobile applications and on-device experiences but can also be a key enabler of privacy and security allowing a user to gain the benefits of neural networks without needing to send their data to the server to be evaluated. This work addresses the challenge of achieving a good trade-off between accuracy and speed for efficient deployment of deep-learning-based pedestrian detection in smart camera applications. A computationally efficient architecture is introduced based on separable convolutions and proposes integrating dense connections across layers and multi-scale feature fusion to improve representational capacity while decreasing the number of parameters and operations. In particular, the contributions of this work are the following: 1) An efficient backbone combining multi-scale feature operations, 2) a more elaborate loss function for improved localization, 3) an anchor-less approach for detection, The proposed approach called YOLOpeds is evaluated using the PETS2009 surveillance dataset on 320x320 images. Overall, YOLOpeds provides real-time sustained operation of over 30 frames per second with detection rates in the range of 86% outperforming existing deep learning models.
Deep Learning based models are currently dominating most state-of-the-art solutions for disease prediction. Existing works employ RNNs along with multiple levels of attention mechanisms to provide interpretability. These deep learning models, with trainable parameters running into millions, require huge amounts of compute and data to train and deploy. These requirements are sometimes so huge that they render usage of such models as unfeasible. We address these challenges by developing a simpler yet interpretable non-deep learning based model for application to EHR data. We model and showcase our work's results on the task of predicting first occurrence of a diagnosis, often overlooked in existing works. We push the capabilities of a tree based model and come up with a strong baseline for more sophisticated models. Its performance shows an improvement over deep learning based solutions (both, with and without the first-occurrence constraint) all the while maintaining interpretability.
The linear inverse problem is fundamental to the development of various scientific areas. Innumerable attempts have been carried out to solve different variants of the linear inverse problem in different applications. Nowadays, the rapid development of deep learning provides a fresh perspective for solving the linear inverse problem, which has various well-designed network architectures results in state-of-the-art performance in many applications. In this paper, we present a comprehensive survey of the recent progress in the development of deep learning for solving various linear inverse problems. We review how deep learning methods are used in solving different linear inverse problems, and explore the structured neural network architectures that incorporate knowledge used in traditional methods. Furthermore, we identify open challenges and potential future directions along this research line.
In this work we present WGANVO, a Deep Learning based monocular Visual Odometry method. In particular, a neural network is trained to regress a pose estimate from an image pair. The training is performed using a semi-supervised approach. Unlike geometry based monocular methods, the proposed method can recover the absolute scale of the scene without neither prior knowledge nor extra information. The evaluation of the system is carried out on the well-known KITTI dataset where it is shown to work in real time and the accuracy obtained is encouraging to continue the development of Deep Learning based methods.
Mobile microscopy is a promising technology to assist and to accelerate disease diagnostics, with its widespread adoption being hindered by the mediocre quality of acquired images. Although some paired image translation and super-resolution approaches for mobile microscopy have emerged, a set of essential challenges, necessary for automating it in a high-throughput setting, still await to be addressed. The issues like in-focus/out-of-focus classification, fast scanning deblurring, focus-stacking, etc. -- all have specific peculiarities when the data are recorded using a mobile device. In this work, we aspire to create a comprehensive pipeline by connecting a set of methods purposely tuned to mobile microscopy: (1) a CNN model for stable in-focus / out-of-focus classification, (2) modified DeblurGAN architecture for image deblurring, (3) FuseGAN model for combining in-focus parts from multiple images to boost the detail. We discuss the limitations of the existing solutions developed for professional clinical microscopes, propose corresponding improvements, and compare to the other state-of-the-art mobile analytics solutions.
In recent years, machine learning and deep learning have become popular methods for financial data analysis, including financial textual data, numerical data, and graphical data. This paper proposes to use sentiment analysis to extract useful information from multiple textual data sources and a blending ensemble deep learning model to predict future stock movement. The blending ensemble model contains two levels. The first level contains two Recurrent Neural Networks (RNNs), one Long-Short Term Memory network (LSTM) and one Gated Recurrent Units network (GRU), followed by a fully connected neural network as the second level model. The RNNs, LSTM, and GRU models can effectively capture the time-series events in the input data, and the fully connected neural network is used to ensemble several individual prediction results to further improve the prediction accuracy. The purpose of this work is to explain our design philosophy and show that ensemble deep learning technologies can truly predict future stock price trends more effectively and can better assist investors in making the right investment decision than other traditional methods.
Data-driven model interpretability is a requirement to gain the acceptance of process engineers to rely on the prediction of a data-driven model to regulate industrial processes in the ironmaking industry. In the research presented in this paper, we focus on the development of an interpretable multivariate time series forecasting deep learning architecture for the temperature of the hot metal produced by a blast furnace. A Long Short-Term Memory (LSTM) based architecture enhanced with attention mechanism and guided backpropagation is proposed to accommodate the prediction with a local temporal interpretability for each input. Results are showing high potential for this architecture applied to blast furnace data and providing interpretability correctly reflecting the true complex variables relations dictated by the inherent blast furnace process, and with reduced prediction error compared to a recurrent-based deep learning architecture.
The global volume of digital data is expected to reach 175 zettabytes by 2025. The volume, variety, and velocity of water-related data are increasing due to large-scale sensor networks and increased attention to topics such as disaster response, water resources management, and climate change. Combined with the growing availability of computational resources and popularity of deep learning, these data are transformed into actionable and practical knowledge, revolutionizing the water industry. In this article, a systematic review of literature is conducted to identify existing research which incorporates deep learning methods in the water sector, with regard to monitoring, management, governance and communication of water resources. The study provides a comprehensive review of state-of-the-art deep learning approaches used in the water industry for generation, prediction, enhancement, and classification tasks, and serves as a guide for how to utilize available deep learning methods for future water resources challenges. Key issues and challenges in the application of these techniques in the water domain are discussed, including the ethics of these technologies for decision-making in water resources management and governance. Finally, we provide recommendations and future directions for the application of deep learning models in hydrology and water resources.
In this thesis, we disentangle the generalized Gauss-Newton and approximate inference for Bayesian deep learning. The generalized Gauss-Newton method is an optimization method that is used in several popular Bayesian deep learning algorithms. Algorithms that combine the Gauss-Newton method with the Laplace and Gaussian variational approximation have recently led to state-of-the-art results in Bayesian deep learning. While the Laplace and Gaussian variational approximation have been studied extensively, their interplay with the Gauss-Newton method remains unclear. Recent criticism of priors and posterior approximations in Bayesian deep learning further urges the need for a deeper understanding of practical algorithms. The individual analysis of the Gauss-Newton method and Laplace and Gaussian variational approximations for neural networks provides both theoretical insight and new practical algorithms. We find that the Gauss-Newton method simplifies the underlying probabilistic model significantly. In particular, the combination of the Gauss-Newton method with approximate inference can be cast as inference in a linear or Gaussian process model. The Laplace and Gaussian variational approximation can subsequently provide a posterior approximation to these simplified models. This new disentangled understanding of recent Bayesian deep learning algorithms also leads to new methods: first, the connection to Gaussian processes enables new function-space inference algorithms. Second, we present a marginal likelihood approximation of the underlying probabilistic model to tune neural network hyperparameters. Finally, the identified underlying models lead to different methods to compute predictive distributions. In fact, we find that these prediction methods for Bayesian neural networks often work better than the default choice and solve a common issue with the Laplace approximation.