On Data Augmentation for GAN Training Tran, Ngoc-Trung; Tran, Viet-Hung; Nguyen, Ngoc-Bao ...
IEEE transactions on image processing,
2021, Letnik:
30
Journal Article
Recenzirano
Odprti dostop
Recent successes in Generative Adversarial Networks (GAN) have affirmed the importance of using more data in GAN training. Yet it is expensive to collect data in many domains such as medical ...applications. Data Augmentation (DA) has been applied in these applications. In this work, we first argue that the classical DA approach could mislead the generator to learn the distribution of the augmented data, which could be different from that of the original data. We then propose a principled framework, termed Data Augmentation Optimized for GAN (DAG), to enable the use of augmented data in GAN training to improve the learning of the original distribution. We provide theoretical analysis to show that using our proposed DAG aligns with the original GAN in minimizing the Jensen-Shannon (JS) divergence between the original distribution and model distribution. Importantly, the proposed DAG effectively leverages the augmented data to improve the learning of discriminator and generator. We conduct experiments to apply DAG to different GAN models: unconditional GAN, conditional GAN, self-supervised GAN and CycleGAN using datasets of natural images and medical images. The results show that DAG achieves consistent and considerable improvements across these models. Furthermore, when DAG is used in some GAN models, the system establishes state-of-the-art Fréchet Inception Distance (FID) scores. Our code is available ( https://github.com/tntrung/dag-gans ).
Respiration signals are a vital sign of life. Monitoring human breath provides critical information for health assessment, diagnosis, and treatment for respiratory diseases such as asthma, chronic ...bronchitis, and emphysema. Stretchable and wearable respiration sensors have recently attracted considerable interest toward monitoring physiological signals in the era of real time and portable healthcare systems. This review provides a snapshot on the recent development of stretchable sensors and wearable technologies for respiration monitoring. The article offers the fundamental guideline on the sensing mechanisms and design concepts of stretchable sensors for detecting vital breath signals such as temperature, humidity, airflow, stress and strain. A highlight on the recent progress in the integration of variable sensing components outlines feasible pathways towards multifunctional and multimodal sensor platforms. Structural designs of nanomaterials and platforms for stretchable respiration sensors are reviewed.
•We review the sensing mechanisms and concepts of stretchable respiration sensors.•We provide detailed integration of multifunctional physiological platforms.•We summarize the development of advanced materials for respiration sensors.
In this article, we propose a novel algorithm, namely PETRELS-ADMM, to deal with subspace tracking in the presence of outliers and missing data. The proposed approach consists of two main stages: ...outlier rejection and subspace estimation. In the first stage, alternating direction method of multipliers (ADMM) is effectively exploited to detect outliers affecting the observed data. In the second stage, we propose an improved version of the parallel estimation and tracking by recursive least squares (PETRELS) algorithm to update the underlying subspace in the missing data context. We then present a theoretical convergence analysis of PETRELS-ADMM which shows that it generates a sequence of subspace solutions converging to the optimum of its batch counterpart. The effectiveness of the proposed algorithm, as compared to state-of-the-art algorithms, is illustrated on both simulated and real data.
Stretchable and wearable sensor technology has attracted significant interests and created high technological impact on portable healthcare and smart human–machine interfaces. Wearable ...electromechanical systems are an important part of this technology that has recently witnessed tremendous progress toward high‐performance devices for commercialization. Over the past few years, great attention has been paid to simultaneously enhance the sensitivity and stretchability of the electromechanical sensors toward high sensitivity, ultra‐stretchability, low power consumption or self‐power functionalities, miniaturisation as well as simplicity in design and fabrication. This work presents state‐of‐the‐art advanced materials and rational designs of electromechanical sensors for wearable applications. Advances in various sensing concepts and structural designs for intrinsic stretchable conductive materials as well as advanced rational platforms are discussed. In addition, the practical applications and challenges in the development of stretchable electromechanical sensors are briefly mentioned and highlighted.
Stretchable electromechanical sensors that can monitor physical and mechanical stimuli generated by the human body are of considerable interest. The design of eletromechanical sensors based on appropriate rational structures and materials will enable the development of ultrasensitive and ultrastretchable sensing devices for human‐activity monitoring, portable healthcare, and smart human–machine interfaces.
Digital polymerase chain reaction (dPCR) technology has remained a "hot topic" in the last two decades due to its potential applications in cell biology, genetic engineering, and medical diagnostics. ...Various advanced techniques have been reported on sample dispersion, thermal cycling and output monitoring of digital PCR. However, a fully automated, low-cost and handheld digital PCR platform has not been reported in the literature. This paper attempts to critically evaluate the recent developments in techniques for sample dispersion, thermal cycling and output evaluation for dPCR. The techniques are discussed in terms of hardware simplicity, portability, cost-effectiveness and suitability for automation. The present paper also discusses the research gaps observed in each step of dPCR and concludes with possible improvements toward portable, low-cost and automatic digital PCR systems.
Silicon carbide (SiC) is one of the most promising materials for applications in harsh environments thanks to its excellent electrical, mechanical, and chemical properties. The piezoresistive effect ...of SiC has recently attracted a great deal of interest for sensing devices in hostile conditions. This paper reviews the piezoresistive effect of SiC for mechanical sensors used at elevated temperatures. We present experimental results of the gauge factors obtained for various poly-types of SiC films and SiC nanowires, the related theoretical analysis, and an overview on the development of SiC piezoresistive transducers. The review also discusses the current issues and the potential applications of the piezoresistive effect in SiC.
Epilepsy is one of the most common brain disorders. For epilepsy diagnosis or treatment, the neurologist needs to observe epileptic spikes from electroencephalography (EEG) data. Since multi-channel ...EEG records can be naturally represented by multi-way tensors, it is of interest to see whether tensor decomposition is able to analyze EEG epileptic spikes.
In this paper, we first proposed the problem of simultaneous multilinear low-rank approximation of tensors (SMLRAT) and proved that SMLRAT can obtain local optimum solutions by using two well-known tensor decomposition algorithms (HOSVD and Tucker-ALS). Second, we presented a new system for automatic epileptic spike detection based on SMLRAT.
We propose to formulate the problem of feature extraction from a set of EEG segments, represented by tensors, as the SMLRAT problem. Efficient EEG features were obtained, based on estimating the 'eigenspikes' derived from nonnegative GSMLRAT. We compared the proposed tensor analysis method with other common tensor methods in analyzing EEG signal and compared the proposed feature extraction method with the state-of-the-art methods. Experimental results indicated that our proposed method is able to detect epileptic spikes with high accuracy.
Our method, for the first time, makes a step forward for automatic detection EEG epileptic spikes based on tensor decomposition. The method can provide a practical solution to distinguish epileptic spikes from artifacts in real-life EEG datasets.
Display omitted
•The runner-root algorithm (RRA) is adapted to solve the network reconfiguration problem.•Five objectives namely power loss, load balancing among the branches, load balancing among ...the feeders, number of switching operations and node voltage deviation are considered.•The proposed RRA method is applied to the 33-bus and 70-bus test networks for evaluation.•The proposed RRA method has better performance in comparison to other methods.
This paper presents a runner-root algorithm (RRA) for electric distribution network reconfiguration (NR) problem. The considered NR problem in this paper is to minimize real power loss, load balancing among the branches, load balancing among the feeders as well as number of switching operations and node voltage deviation using max-min method for selection of the final compromised solution. RRA is equipped with two explorative tools, which are random jumps with large steps and re-initialization strategy to escape from local optimal. Moreover, RRA is also equipped with an exploitative tool to search around the current best solution with large and small steps to ensure the obtained result of global optimization. The effectiveness of the applied RRA in both single- and multi-objective has been tested on 33-node and 70-node distribution network systems and the obtained test results have been compared to those from other methods in the literature. The simulation results show that the applied RRA can be an efficient method for network reconfiguration problems with single- and multi-objective.
In this study, cobalt ferrite coated carbon felt (CoFe2O4/CF) was synthesized by solvothermal method and applied as cathode for electro-Fenton (EF) treatment of tartrazine (TTZ) in water. The ...materials were characterized by SEM, XRD, FTIR, CV, and EIS to explore their physical, chemical, and electrical properties. The effects of solvothermal temperature and metal content on the TTZ removal were examined, showing that 220 °C with 2 mM of Co and 4 mM of Fe precursors were the best synthesis condition. Various influencing factors such as applied current density, pH, TTZ concentration, and electrolytes were investigated, and the optimal condition was found at 8.33 mA cm−2, pH 3, 50 mgTTZ L−1, and 50 mM of Na2SO4, respectively. By radical quenching test, ▪, 1O2, and HO were recognized as the key reactive oxygen species and the reaction mechanism was proposed for the EF decolorization of TTZ using CoFe2O4/CF cathode. The reusability and stability test showed that the highly efficient CoFe2O4/CF cathode is very promising for practical application in wastewater treatment, especially for dyes and other recalcitrant organic compounds to improve its biodegradability.
•CoFe2O4/CF was synthesized as an effective cathode for electro-Fenton reaction.•Systematic study with different material synthesis and TTZ reaction conditions.•▪1O2, and HO as the key reactive oxygen species via radical quenching test.•Reaction mechanism was proposed for the electro-Fenton using CoFe2O4/CF cathode.
Microelectromechanical systems sensors have been intensively developed utilizing various physical concepts, such as piezoresistive, piezoelectric, and thermoresistive effects. Among these sensing ...concepts, the thermoresistive effect is of interest for a wide range of thermal sensors and devices, thanks to its simplicity in implementation and high sensitivity. The effect of temperature on the electrical resistance of some metals and semiconductors has been thoroughly investigated, leading to the significant growth and successful demonstration of thermal-based sensors, such as temperature sensors, convective accelerometers and gyroscopes, and thermal flow sensors. In this paper, we review the fundamentals of the thermoresistive effect in metals and semiconductors. We also discuss the influence of design and fabrication parameters on the thermoresistive sensitivity. This paper includes several desirable features of thermoresistive sensors and recent developments in these sensors are summarized. This review provides insights into how it is affected by various parameters, and useful guidance for industrial designers in terms of high sensitivity and linearity and fast response.