Regional dropout strategies have been proposed to enhance the performance of convolutional neural network classifiers. They have proved to be effective for guiding the model to attend on less discriminative parts of objects (e.g. leg as opposed to head of a person), thereby letting the network generalize better and have better object localization capabilities. On the other hand, current methods for regional dropout removes informative pixels on training images by overlaying a patch of either black pixels or random noise. Such removal is not desirable because it leads to information loss and inefficiency during training. We therefore propose the CutMix augmentation strategy: patches are cut and pasted among training images where the ground truth labels are also mixed proportionally to the area of the patches. By making efficient use of training pixels and retaining the regularization effect of regional dropout, CutMix consistently outperforms the stateof-the-art augmentation strategies on CIFAR and ImageNet classification tasks, as well as on the ImageNet weaklysupervised localization task. Moreover, unlike previous augmentation methods, our CutMix-trained ImageNet classifier, when used as a pretrained model, results in consistent performance gains in Pascal detection and MS-COCO image captioning benchmarks. We also show that CutMix improves the model robustness against input corruptions and its out-of-distribution detection performances.

Scene text detection methods based on neural networks have emerged recently and have shown promising results. Previous methods trained with rigid word-level bounding boxes exhibit limitations in representing the text region in an arbitrary shape. In this paper, we propose a new scene text detection method to effectively detect text area by exploring each character and affinity between characters. To overcome the lack of individual character level annotations, our proposed framework exploits both the given characterlevel annotations for synthetic images and the estimated character-level ground-truths for real images acquired by the learned interim model. In order to estimate affinity between characters, the network is trained with the newly proposed representation for affinity. Extensive experiments on six benchmarks, including the TotalText and CTW-1500 datasets which contain highly curved texts in natural images, demonstrate that our character-level text detection significantly outperforms the state-of-the-art detectors. According to the results, our proposed method guarantees high flexibility in detecting complicated scene text images, such as arbitrarily-oriented, curved, or deformed texts.

Many new proposals for scene text recognition (STR) models have been introduced in recent years. While each claim to have pushed the boundary of the technology, a holistic and fair comparison has been largely missing in the field due to the inconsistent choices of training and evaluation datasets. This paper addresses this difficulty with three major contributions. First, we examine the inconsistencies of training and evaluation datasets, and the performance gap results from inconsistencies. Second, we introduce a unified four-stage STR framework that most existing STR models fit into. Using this framework allows for the extensive evaluation of previously proposed STR modules and the discovery of previously unexplored module combinations. Third, we analyze the module-wise contributions to performance in terms of accuracy, speed, and memory demand, under one consistent set of training and evaluation datasets. Such analyses clean up the hindrance on the current comparisons to understand the performance gain of the existing modules. Our code is publicly available 1 .

Deep convolutional neural networks (DCNNs) have shown remarkable performance in image classification tasks

SignificanceThe optical method to determine oxygen saturation in blood is limited to only tissues that can be transilluminated. The status quo provides a single-point measurement and lacks 2D oxygenation mapping capability. We use organic printed optoelectronics in a flexible array configuration that senses reflected light from tissue. Our reflectance oximeter is used beyond conventional sensing locations and accurately measures oxygen saturation on the forehead. In a full system implementation, coupled with a mathematical model, we create 2D oxygenation maps of adult forearms under pressure-cuff–induced ischemia. Our skin-like flexible sensor system has the potential to transform oxygenation monitoring of tissues, wounds, skin grafts, and transplanted organs.

The fast development of the Internet of Things (IoT) technology in recent years has supported connections of numerous smart things along with sensors and established seamless data exchange between them, so it leads to a stringy requirement for data analysis and data storage platform such as cloud computing and fog computing. Healthcare is one of the application domains in IoT that draws enormous interest from industry, the research community, and the public sector. The development of IoT and cloud computing is improving patient safety, staff satisfaction, and operational efficiency in the medical industry. This survey is conducted to analyze the latest IoT components, applications, and market trends of IoT in healthcare, as well as study current development in IoT and cloud computing-based healthcare applications since 2015. We also consider how promising technologies such as cloud computing, ambient assisted living, big data, and wearables are being applied in the healthcare industry and discover various IoT, e-health regulations and policies worldwide to determine how they assist the sustainable development of IoT and cloud computing in the healthcare industry. Moreover, an in-depth review of IoT privacy and security issues, including potential threats, attack types, and security setups from a healthcare viewpoint is conducted. Finally, this paper analyzes previous well-known security models to deal with security risks and provides trends, highlighted opportunities, and challenges for the IoT-based healthcare future development.

The interfacing of soft and hard electronics is a key challenge for flexible hybrid electronics. Currently, a multisubstrate approach is employed, where soft and hard devices are fabricated or assembled on separate substrates, and bonded or interfaced using connectors; this hinders the flexibility of the device and is prone to interconnect issues. Here, a single substrate interfacing approach is reported, where soft devices, i.e., sensors, are directly printed on Kapton polyimide substrates that are widely used for fabricating flexible printed circuit boards (FPCBs). Utilizing a process flow compatible with the FPCB assembly process, a wearable sensor patch is fabricated composed of inkjet‐printed gold electrocardiography (ECG) electrodes and a stencil‐printed nickel oxide thermistor. The ECG electrodes provide 1 mVp–p ECG signal at 4.7 cm electrode spacing and the thermistor is highly sensitive at normal body temperatures, and demonstrates temperature coefficient, α ≈ –5.84% K–1 and material constant, β ≈ 4330 K. This sensor platform can be extended to a more sophisticated multisensor platform where sensors fabricated using solution processable functional inks can be interfaced to hard electronics for health and performance monitoring, as well as internet of things applications.

In this paper, we introduce a novel technique to automatically detect salient regions of an image via highdimensional color transform. Our main idea is to represent a saliency map of an image as a linear combination of high-dimensional color space where salient regions and backgrounds can be distinctively separated. This is based on an observation that salient regions often have distinctive colors compared to the background in human perception, but human perception is often complicated and highly nonlinear. By mapping a low dimensional RGB color to a feature vector in a high-dimensional color space, we show that we can linearly separate the salient regions from the background by finding an optimal linear combination of color coefficients in the high-dimensional color space. Our high dimensional color space incorporates multiple color representations including RGB, CIELab, HSV and with gamma corrections to enrich its representative power. Our experimental results on three benchmark datasets show that our technique is effective, and it is computationally efficient in comparison to previous state-of-the-art techniques.