We present a target-agnostic adversarial autoencoder with conditional normalizing flows specifically designed to, given any unlabeled image dataset, purify adversarial samples into clean images, i.e., remove adversarial noise from the images while preserving their visual quality. In our model interpretation, samples are processed by manifold projection in which the encoder brings the sample back into a posterior data distribution in latent space so that the sample is less likely to be irregular to the learned representation of any target classifier. Normalizing flows conditioned on top of our hybrid network structure and walk-back training are used to deal with common drawbacks of generative model and autoencoder-based approaches: not only the trade-off between compression loss and over-fitting on training data but also the structural model dependency on dataset classes and labels. Experiments demonstrated that our proposed model is preferable to existing target-agnostic adversarial defense methods particularly for large and unlabeled image datasets.

With deep neural networks based function approximators, reinforcement learning holds the promise of learning complex end-to-end robotic controllers that can map high-dimensional sensory information directly to control policies. However, a common challenge, especially for robotics, is sample-efficient learning from sparse rewards, in which an agent is required to find a long sequence of “correct” actions to achieve a desired outcome. Unfortunately, inevitable perturbations on observations may make this task trickier to solve. Here, this paper advances a novel robust goal-conditioned reinforcement learning approach for end-to-end robotic control in adversarial and sparse reward environments. Specifically, a mixed adversarial attack scheme is presented to generate diverse adversarial perturbations on observations by combining white-box and black-box attacks. Meanwhile, a hindsight experience replay technique considering observation perturbations is developed to turn a failed experience into a successful one and generate the policy trajectories perturbed by the mixed adversarial attacks. Additionally, a robust goal-conditioned actor-critic method is proposed to learn goal-conditioned policies and keep the variations of the perturbed policy trajectories within bounds. Finally, the proposed method is evaluated on three tasks with adversarial attacks and sparse reward settings. The results indicate that our scheme can ensure robotic control performance and policy robustness on the adversarial and sparse reward tasks.

Uncertainty estimation is a critical component of building safe and reliable machine learning models. Accurate estimation of uncertainties is essential for identifying and mitigating potential risks and ensuring that machine learning systems operate reliably in real-world scenarios. Various approaches, such as ensemble and Bayesian neural networks have been developed by sampling probability predictions from submodels, which is computatinally expensive. Currently, these methods are unable to clearly define the boundary between in distribution (ID) and out-of-distribution (OOD) data. To fill up this research gap, this paper presents a normalizing flow based framework to directly predict parameters of prior distributions over the probability with a neural network, which is capable of effectively distinguishing ID and OOD data for regression problems. The posterior distributions learned by the proposed model accurately model uncertainties for OOD data from ID data without requiring OOD data at training time. This approach has shown promising results in a number of applications, including image depth estimation and image adversarial attacks.

This research develops a methodology to identify transactions through data-driven tracking and analysis of ransomware-Bitcoin payment networks [30]. We demonstrate the methodology by applying the GraphSAGE embedding algorithm to the WannaCry ransomware-Bitcoin cash-out network. The paper takes a data-driven approach to building a machine learning system that allows analysts to define features relevant to ransomware-Bitcoin payment networks.

This paper developed an ultrasound concave 2D ring array transducer for retinal visual stimulation of the brain to restore vision in blind people. Theory, simulation, and experiment were presented.Â

Impedance-based method has been increasingly adopted to assess the stability of inverter-based resources (IBRs). To get a better view of the state-of-art and challenges for implementing the impedance-based dynamic analysis, a survey with general/specific questions has been initiated by IEEE Task Force on Frequency-Domain Modeling and Dynamic Analysis of HVDC and FACTS. The feedbacks are collected from universities, national labs, transmission system operators (TSOs), power plant developers, as well as IBR vendors. It is interesting to note that while many common understandings have been established in practices, certain gaps still exist among different stakeholders. This article intends to bridge this gap by sharing a summary of the survey, including questionnaires, responses from different stakeholders, and the analysis of survey results. The challenges for different stakeholders using impedance-based method are identified, which hopefully shed a light on the future research work.

Sound event location is a momentous subtask of two-dimensional direction-of-arrival (2D-DOA) estimation to forecast the azimuth and elevation angle from each active sound event class of an audio fragment using multi-label regression in the 3D Cartesian coordinates. The main problem with the traditional multi-signal classification (MUSIC) algorithm and the existing baseline convolution recurrent neural network (BCRNN) are lower precision and huge computation in weaker signal-noise-ratio (SNR) environment. In particular, MUSIC algorithm will cause complete distortion when the SNR is lower than -5 dB. We thus design effective residual self-attention recurrent neural network (ESRNN) not only to overcome the distortion of traditional MUSIC algorithm under lower SNR but also further reduce predicted 2D-DOA error in different SNR reverberation environments. Two different filter structures, ESRNN-L and ESRNN-G, are designed to improve the SNR of our model when the SNR is above 0 dB or below -5 dB in a targeted manner respectively. To verify the efficiency and robustness, we train and test our model using TAU Spatial Sound Events 2019 datasets extracted the phase and magnitude spectrogram with different synthetic SNRs ranging from -10 dB to 30 dB. The experimental results show that the optimal 2D-DOA error of our ESRNN-L is 21% lower than BCRNN when the SNR is less than -5 dB, and ESRNN-G is 15% error lower with 10% parameters reduction when the SNR is higher than 0 dB.

This work introduces an open-source hardware and software expansion for digital microfluidics (DMF) devices that integrates programmatic temperature and magnetic control. Our aim was to create a versatile, modular, extension that may aid in biomedical applications and has proven potential for use in bacteriophage research. To achieve this, we developed an accessible DMF device extension called PhageBox. PhageBox incorporates temperature and magnetic control modules at the hardware level. At the software level, it is controlled by embedded software that includes a unique model for bio-protocol programming and a graphical user interface for visual feedback and operation of the device. We evaluated PhageBox’s effectiveness in biomedical applications using functional testing and validated the temperature control using thermography, achieving a range of +/-0.2 Celsius. The electromagnets produced a magnetic force of 15 milliTesla, which precisely immobilized magnetic beads. Furthermore, we demonstrate the potential of PhageBox in bacteriophage research through three initial proof of principle protocols: (1) a universal framework for PCR, (2) T7 bacteriophage restriction enzyme digestion, and (3) the concentration of PhiX174 RF genomic DNA. Overall, our study presents an open-source solution for DMF devices to integrate temperature regulation and magnetism, facilitating various applications in biomedical research and bacteriophage studies. (GitHub: https://github.com/Dreycey/PhageBox)

This letter derives and discusses the superiority of a simple dc-link capacitor voltage control configuration for multilevel neutral-point-clamped converters with any number of levels. The control involves n−2 control loops regulating the difference between the voltage of neighbor capacitors. These control loops are inherently decoupled; i.e., they are independent and the control action of one loop does not affect the others. The good performance of such control is confirmed through simulations and experiments.

Line waves are modes that propagate along a line connecting two-dimensional impedance surfaces with dual electromagnetic responses. In general, line waves can be implemented using complementary impedances or impedances with a gain-loss effect. On the other hand, if the two impedance surfaces are non-complementary, they can support a different type of mode called a “quasi-line wave”. We propose a different configuration to implement quasi-line waves. Indeed, non-dual structures with purely inductive impedances are used in order to drive the mode. We investigate the implementation of quasi-line modes using multilayer graphenes. This mode has a wide bandwidth in the terahertz range and a significant propagation length. The proposed structure guides one-way modes in which the fields are highly concentrated around the edges of two inductive impedance structures. Furthermore, incorporating graphene into the proposed waveguide allows for control of bandwidth and transmission characteristic of the waveguide.

A generative adversarial network (GAN) enables mapping a data sample from one domain to another one. It has extensively been employed in image-to-image and text-toimage translation. We propose an EEG-to-EEG translation model based on a GAN structure in which a conditional GAN (cGAN) is combined with a variational autoencoder (VAE), named as VAE-cGAN. Scalp EEG (sEEG) is plagued by noise and suffers from low resolution. On the other hand, intracranial EEG (iEEG) recordings enjoy high resolution. Here, we consider the task of mapping the sEEG to iEEG to enhance the sEEG resolution.  

In this work, we propose a pseudospin-polarized closed waveguide whose walls are formed by dual electromagnetic boundary conditions.  The structure has a very wide bandwidth so that it covers an extremely wide electromagnetic spectrum.  The proposed waveguide supports one-way pseudospin states resulting from the establishment of dual relationships at the waveguide boundaries, and hence the establishment of mirror reflection symmetry. This provides the proposed waveguide with the ability to suppress backscattering.  The boundary conditions of the proposed structure consist of complementary metasurfaces which have dual surface impedances.  The presence of these boundary conditions leads to forming a pseudospin-polarized waveguide having the ability to suppress and filter the backward modes that may be excited along bent paths or discontinuities.  To the best of our knowledge, this is the widest bandwidth waveguide that has been proposed so far, so that the isolation of the waveguide in the frequency range of 7- 350 GHz is above -0.5 dB.  In addition, we use a very precise method based on the variational method to calculate the surface impedance of the metasurfaces forming the waveguide boundaries so as to increase the accuracy of the obtained results.

Measurements of the average properties of vehicular traffic are inherently noisy. The distributions of flow and speed measurements at any particular density are non-Gaussian with density-dependent variance, skewness, and kurtosis. Previous studies have failed to properly account for these complicated noise properties. In remediation, we present FitFun, a general framework for modelling any observed flow-density-speed relationship. Models specified within FitFun incorporate components for both the functional form and the noise. We define three flexible noise model components and we fit 200 different models to a high-quality sample of 10,150 observed urban flow-occupancy relationships. We compare the fits using information criteria and assess fit quality through analysis of the residuals. We find that the non-parametric Sun model for the functional form component combined with a Skew Exponential Power Type III noise component significantly outperforms all of the other models. Interestingly, we find that the city, country, road topology, and detector location have virtually no impact on model performance and fit quality, which is very convenient for model selection. The only factor of relevance from those that we studied is the effective occupancy coverage of the data. We conclude that certain models specified judiciously within FitFun can successfully capture the functional form and noise of observed flow-density-speed relationships without the need to discard data taken during non-stationary conditions. This is particularly advantageous for urban data where stationary traffic conditions are rarely observed. Accepted by Transportation Research Part C on 16th Feb 2023.

In this work, we propose and implement a bijunction depletion-type silicon electro-optic phase shifter. The phase shifter has a lateral profile of implants, that closely resembles that of a common bipolar junction transistor, and thus, has two polarities. These are acceptor-donor-acceptor (PNP) as well as donor-acceptor-donor (NPN). We realize both variants in IMEC ISIPP50G open-access silicon photonic technology and compare them to lateral and interleaved phase shifters. Both PNP and NPN phase shifters exhibit a VπLπ figure that is at least 14.47% and up to 45.1% lower than that of the lateral and interleaved phase shifters realized in the same technology. Bi-junction phase shifters can be implemented in any planar silicon photonic technology that offers bipolar implantations within silicon photonic waveguides.

In this paper, we validate the proposed model by comparing the distributions of the model parameters, extracted from the simulated channels, with their counterparts extracted from the measured channels. The validation shows that the proposed model provides a very good representation for the V2VRU propagation channel in the considered scenarios. This channel model can be used in simulations to develop and evaluate V2VRU communication and collision avoidance algorithms in critical accident scenarios.

As the echoes of ChatGPT’s remarkable success continue to permeate the AI community, its formidable successor, GPT-4, has emerged, showing off a wealth of novel features. In this concise paper, we elucidate the capabilities of GPT-4 and conduct a comparative analysis with its predecessor, ChatGPT, offering insights into their relative strengths and advancements in the rapidly evolving field of generative AI. We also present a comprehensive summary of the major performance results reported by OpenAI on GPT-4 across various Natural Language Processing (NLP) tasks. We place great emphasis on the innovative advancements offered by GPT-4 in comparison to its predecessors. Our focus is on highlighting its remarkable performance while also mentioning its limitations. The purpose of this paper is to deliver a succinct understanding of the new features and performance benchmarks of GPT-4.

Non-fungible tokens (NFTs) have been attracting the interest of both technical and non-technical parties, including collectors and traders, among others. The number of transactions in NFTs surpassed $50 billion in 2022. Blockchain technology’s advantages as a distributed, immutable, and transparent database make it ideal for verifying the ownership of digital goods created by their producers. On the other hand, high computation and transaction costs are known disadvantages of public blockchain networks like Ethereum V1, used in NFT marketplaces. To address these inefficiencies, other public blockchain systems have emerged as replacements for NFT marketplaces, each with its own unique properties. When planning such an NFT exchange, it is vital, but not trivial, to select the most appropriate public blockchain platform. In this work, we make two contributions to support this decision. In this paper, we present IntelliChain, a self-adaptive framework that can predict the best optimal transaction fees (also known as a gas fee) for blockchain to reduce the errors and also an ability to switch the public blockchain-based dynamic needs such as transaction fees and stability of the network.

This study delves into the capabilities of reconfig- urable intelligent surfaces (RISs) in enhancing bidirectional non- orthogonal multiple access (NOMA) networks. The proposed approach partitions RIS to optimize the channel conditions for NOMA users, improving NOMA gain and eliminating the re- quirement for uplink (UL) power control. The proposed approach is rigorously evaluated under four practical operational regimes; 1) Quality-of-Service (QoS) sufficient regime, 2) RIS and power efficient regime, 3) max-min fair regime, and 4) maximum throughput regime, each subject to both UL and downlink (DL) QoS constraints. By leveraging decoupled nature of RIS portions and base station (BS) transmit power, closed-form solutions are derived to demonstrate how optimal RIS partitioning can meet UL-QoS requirements while optimal BS power control can ensure DL-QoS compliance. Our analytical findings are validated through simulations, highlighting the significant benefits that RISs can bring to the NOMA networks in the aforementioned operational scenarios.