Search

Fluid antennas present a relatively new idea for harnessing the fading and interference issues in multiple user wireless systems, such as 6G. Here, we systematically compare their unique radiation beam forming mechanism to the existing multiple-antenna systems in a wireless system. Subsequently, a unified mathematical model for fluid antennas is deduced based on the eigenmode theory. As mathematically derived from the multimode resonant theory, the spectral expansion model of any antennas which occupy variable spaces and have changeable feeding schemes can be generalized as fluid antennas. Non-liquid and liquid fluid antenna examples are presented, simulated and discussed. The symmetry or modal parity of eigenmodes is explored as an additional degree of freedom to design the fluid antennas for future wireless systems. As conceptually deduced and illustrated, the multi-dimensional and continuously adaptive ability of eigenmodes can be considered as the most fundamental intrinsic characteristic of the fluid antenna systems. It opens an uncharted area in the developments of intelligent antennas (IAs), which brings more flexibility to on-demand antenna beam null manipulating techniques for future wireless applications.

Electroencephalogram (EEG) signals have attracted significant attention from researchers due to their non-invasive nature and high temporal sensitivity in decoding visual stimuli. However, most recent studies have focused solely on the relationship between EEG and image data pairs, neglecting the valuable ``beyond-image-modality" information embedded in EEG signals. This results in the loss of critical multimodal information in EEG. To address this limitation, we propose CognitionCapturer, a unified framework that fully leverages multimodal data to represent EEG signals. Specifically, CognitionCapturer trains Modality Expert Encoders for each modality to extract cross-modal information from the EEG modality. Then, it introduces a diffusion prior to map the EEG embedding space to the CLIP embedding space, followed by using a pretrained generative model, the proposed framework can reconstruct visual stimuli with high semantic and structural fidelity. Notably, the framework does not require any fine-tuning of the generative models and can be extended to incorporate more modalities. Through extensive experiments, we demonstrate that CognitionCapturer outperforms state-of-the-art methods both qualitatively and quantitatively. Code: https://github.com/XiaoZhangYES/CognitionCapturer.

Transformer models have gained significant attention due to their power in machine learning tasks. Their extensive deployment has raised concerns about the potential leakage of sensitive information during inference. However, when being applied to Transformers, existing approaches based on secure two-party computation (2PC) bring about efficiency limitations in two folds: (1) resource-intensive matrix multiplications in linear layers, and (2) complex non-linear activation functions like $\mathsf{GELU}$ and $\mathsf{Softmax}$. This work presents a new two-party inference framework $\mathsf{Nimbus}$ for Transformer models. For the linear layer, we propose a new 2PC paradigm along with an encoding approach to securely compute matrix multiplications based on an outer-product insight, which achieves $2.9\times \sim 12.5\times$ performance improvements compared to the state-of-the-art (SOTA) protocol. For the non-linear layer, through a new observation of utilizing the input distribution, we propose an approach of low-degree polynomial approximation for $\mathsf{GELU}$ and $\mathsf{Softmax}$, which improves the performance of the SOTA polynomial approximation by $2.9\times \sim 4.0\times$, where the average accuracy loss of our approach is 0.08\% compared to the non-2PC inference without privacy. Compared with the SOTA two-party inference, $\mathsf{Nimbus}$ improves the end-to-end performance of \bert{} inference by $2.7\times \sim 4.7\times$ across different network settings., Comment: Accepted by NIPS 2024

Scaling up quantum dots to two-dimensional (2D) arrays is a crucial step for advancing semiconductor quantum computation. However, maintaining excellent tunability of quantum dot parameters, including both nearest-neighbor and next-nearest-neighbor couplings, during 2D scaling is challenging, particularly for silicon quantum dots due to their relatively small size. Here, we present a highly controllable and interconnected 2D quantum dot array in planar silicon, demonstrating independent control over electron fillings and the tunnel couplings of nearest-neighbor dots. More importantly, we also demonstrate the wide tuning of tunnel couplings between next-nearest-neighbor dots,which plays a crucial role in 2D quantum dot arrays. This excellent tunability enables us to alter the coupling configuration of the array as needed. These results open up the possibility of utilizing silicon quantum dot arrays as versatile platforms for quantum computing and quantum simulation.

Electron spin qubits in silicon are a promising platform for fault-tolerant quantum computing. Low-frequency noise, including nuclear spin fluctuations and charge noise, is a primary factor limiting gate fidelities. Suppressing this noise is crucial for high-fidelity qubit operations. Here, we report on a two-qubit quantum device in natural silicon with universal qubit control, designed to investigate the upper limits of gate fidelities in a non-purified Si/SiGe quantum dot device. By employing advanced device structures, qubit manipulation techniques, and optimization methods, we have achieved single-qubit gate fidelities exceeding 99% and a two-qubit Controlled-Z (CZ) gate fidelity of 91%. Decoupled CZ gates are used to prepare Bell states with a fidelity of 91%, typically exceeding previously reported values in natural silicon devices. These results underscore that even natural silicon has the potential to achieve high-fidelity gate operations, particularly with further optimization methods to suppress low-frequency noise.

In this paper, we investigate the stability equivalence problem for stochastic differential delay equations, the auxiliary stochastic differential equations and their corresponding Euler-Maruyama (EM) methods under $G$-framework. More precisely, for $p\geq 2$, we prove the equivalence of practical exponential stability in $p$-th moment sense among stochastic differential delay equations driven by $G$-Brownian motion ($G$-SDDEs), the auxiliary stochastic differential equations driven by $G$-Brownian motion ($G$-SDEs), and their corresponding Euler-Maruyama methods, provided the delay or the step size is small enough. Thus, we can carry out careful simulations to examine the practical exponential stability of the underlying $G$-SDDE or $G$-SDE under some reasonable assumptions.

Current video coding standards, including H.264/AVC, HEVC, and VVC, employ discrete cosine transform (DCT), discrete sine transform (DST), and secondary to Karhunen-Loeve transforms (KLTs) decorrelate the intra-prediction residuals. However, the efficiency of these transforms in decorrelation can be limited when the signal has a non-smooth and non-periodic structure, such as those occurring in textures with intricate patterns. This paper introduces a novel adaptive separable path graph-based transform (GBT) that can provide better decorrelation than the DCT for intra-predicted texture data. The proposed GBT is learned in an online scenario with sequential K-means clustering, which groups similar blocks during encoding and decoding to adaptively learn the GBT for the current block from previously reconstructed areas with similar characteristics. A signaling overhead is added to the bitstream of each coding block to indicate the usage of the proposed graph-based transform. We assess the performance of this method combined with H.264/AVC intra-coding tools and demonstrate that it can significantly outperform H.264/AVC DCT for intra-predicted texture data., Comment: 5 pages, 4 figures

Strain engineering has quickly emerged as a viable option to modify the electronic, optical and magnetic properties of 2D materials. However, it remains challenging to arbitrarily control the strain. Here we show that by creating atomically-flat surface nanostructures in hexagonal boron nitride, we achieve an arbitrary on-chip control of both the strain distribution and magnitude on high-quality molybdenum disulfide. The phonon and exciton emissions are shown to vary in accordance with our strain field designs, enabling us to write and draw any photoluminescence color image in a single chip. Moreover, our strain engineering offers a powerful means to significantly and controllably alter the strengths and energies of interlayer excitons at room temperature. This method can be easily extended to other material systems and offers a promise for functional excitonic devices., Comment: 8 pages, 5 figures

Deep neural network (DNN) inference based on secure 2-party computation (2PC) can offer cryptographically-secure privacy protection but suffers from orders of magnitude latency overhead due to enormous communication. Previous works heavily rely on a proxy metric of ReLU counts to approximate the communication overhead and focus on reducing the ReLUs to improve the communication efficiency. However, we observe these works achieve limited communication reduction for state-of-the-art (SOTA) 2PC protocols due to the ignorance of other linear and non-linear operations, which now contribute to the majority of communication. In this work, we present CoPriv, a framework that jointly optimizes the 2PC inference protocol and the DNN architecture. CoPriv features a new 2PC protocol for convolution based on Winograd transformation and develops DNN-aware optimization to significantly reduce the inference communication. CoPriv further develops a 2PC-aware network optimization algorithm that is compatible with the proposed protocol and simultaneously reduces the communication for all the linear and non-linear operations. We compare CoPriv with the SOTA 2PC protocol, CrypTFlow2, and demonstrate 2.1x communication reduction for both ResNet-18 and ResNet-32 on CIFAR-100. We also compare CoPriv with SOTA network optimization methods, including SNL, MetaPruning, etc. CoPriv achieves 9.98x and 3.88x online and total communication reduction with a higher accuracy compare to SNL, respectively. CoPriv also achieves 3.87x online communication reduction with more than 3% higher accuracy compared to MetaPruning., Comment: Accepted by NeurIPS 2023

A systematic investigation of spin injection behavior in Au/FM (FM = Fe and Ni) multilayers is performed using the superdiffusive spin transport theory. By exciting the nonmagnetic layer, the laser-induced hot electrons may transfer spin angular momentum into the adjacent ferromagnetic (FM) metals resulting in ultrafast demagnetization or enhancement. We find that these experimental phenomena sensitively depend on the particular interface reflectivity of hot electrons and may reconcile the different observations in experiment. Stimulated by the ultrafast spin currents carried by the hot electrons, we propose the multilayer structures to generate highly spin polarized currents for development of future ultrafast spintronics devices. The spin polarization of the electric currents carried by the hot electrons can be significantly enhanced by the joint effects of bulk and interfacial spin filtering. Meanwhile the intensity of the generated spin current can be optimized by varying the number of repeated stacking units and the thickness of each metallic layer., Comment: 11 pages, 6 figures

With ChatGPT as a representative, tons of companies have began to provide services based on large Transformers models. However, using such a service inevitably leak users' prompts to the model provider. Previous studies have studied secure inference for Transformer models using secure multiparty computation (MPC), where model parameters and clients' prompts are kept secret. Despite this, these frameworks are still limited in terms of model performance, efficiency, and deployment. To address these limitations, we propose framework PUMA to enable fast and secure Transformer model inference. Our framework designs high quality approximations for expensive functions such as GeLU and softmax, and significantly reduce the cost of secure inference while preserving the model performance. Additionally, we design secure Embedding and LayerNorm procedures that faithfully implement the desired functionality without undermining the Transformer architecture. PUMA is about $2\times$ faster than the state-of-the-art framework MPCFORMER(ICLR 2023) and has similar accuracy as plaintext models without fine-tuning (which the previous works failed to achieve). PUMA can even evaluate LLaMA-7B in around 5 minutes to generate 1 token. To our best knowledge, this is the first time that a model with such a parameter size is able to be evaluated under MPC. PUMA has been open-sourced in the Github repository of SecretFlow-SPU.

Life in complex systems, such as cities and organisms, comes to a standstill when global coordination of mass, energy, and information flows is disrupted. Global coordination is no less important in single cells, especially in large oocytes and newly formed embryos, which commonly use fast fluid flows for dynamic reorganization of their cytoplasm. Here, we combine theory, computing, and imaging to investigate such flows in the Drosophila oocyte, where streaming has been proposed to spontaneously arise from hydrodynamic interactions among cortically anchored microtubules loaded with cargo-carrying molecular motors. We use a fast, accurate, and scalable numerical approach to investigate fluid-structure interactions of 1000s of flexible fibers and demonstrate the robust emergence and evolution of cell-spanning vortices, or twisters. Dominated by a rigid body rotation and secondary toroidal components, these flows are likely involved in rapid mixing and transport of ooplasmic components.

Fast radio bursts (FRBs) are highly dispersed millisecond-duration radio bursts, of which the physical origin is still not fully understood. FRB 20201124A is one of the most actively repeating FRBs. In this paper, we present the collection of 1863 burst dynamic spectra of FRB 20201124A measured with the Five-hundred-meter Aperture Spherical radio Telescope (FAST). The current collection, taken from the observation during the FRB active phase from April to June 2021, is the largest burst sample detected in any FRB so far. The standard PSRFITs format is adopted, including dynamic spectra of the burst, and the time information of the dynamic spectra, in addition, mask files help readers to identify the pulse positions are also provided.

Secure multi-party computation (MPC) enables computation directly on encrypted data and protects both data and model privacy in deep learning inference. However, existing neural network architectures, including Vision Transformers (ViTs), are not designed or optimized for MPC and incur significant latency overhead. We observe Softmax accounts for the major latency bottleneck due to a high communication complexity, but can be selectively replaced or linearized without compromising the model accuracy. Hence, in this paper, we propose an MPC-friendly ViT, dubbed MPCViT, to enable accurate yet efficient ViT inference in MPC. Based on a systematic latency and accuracy evaluation of the Softmax attention and other attention variants, we propose a heterogeneous attention optimization space. We also develop a simple yet effective MPC-aware neural architecture search algorithm for fast Pareto optimization. To further boost the inference efficiency, we propose MPCViT+, to jointly optimize the Softmax attention and other network components, including GeLU, matrix multiplication, etc. With extensive experiments, we demonstrate that MPCViT achieves 1.9%, 1.3% and 3.6% higher accuracy with 6.2x, 2.9x and 1.9x latency reduction compared with baseline ViT, MPCFormer and THE-X on the Tiny-ImageNet dataset, respectively. MPCViT+ further achieves a better Pareto front compared with MPCViT. The code and models for evaluation are available at https://github.com/PKU-SEC-Lab/mpcvit., Comment: Accepted by ICCV 2023 conference

Land cover classification is a multi-class segmentation task to classify each pixel into a certain natural or man-made category of the earth surface, such as water, soil, natural vegetation, crops, and human infrastructure. Limited by hardware computational resources and memory capacity, most existing studies preprocessed original remote sensing images by down sampling or cropping them into small patches less than 512*512 pixels before sending them to a deep neural network. However, down sampling images incurs spatial detail loss, renders small segments hard to discriminate, and reverses the spatial resolution progress obtained by decades of years of efforts. Cropping images into small patches causes a loss of long-range context information, and restoring the predicted results to their original size brings extra latency. In response to the above weaknesses, we present an efficient lightweight semantic segmentation network termed MKANet. Aimed at the characteristics of top view high-resolution remote sensing imagery, MKANet utilizes sharing kernels to simultaneously and equally handle ground segments of inconsistent scales, and also employs parallel and shallow architecture to boost inference speed and friendly support image patches more than 10X larger. To enhance boundary and small segments discrimination, we also propose a method that captures category impurity areas, exploits boundary information and exerts an extra penalty on boundaries and small segment misjudgment. Both visual interpretations and quantitative metrics of extensive experiments demonstrate that MKANet acquires state-of-the-art accuracy on two land-cover classification datasets and infers 2X faster than other competitive lightweight networks. All these merits highlight the potential of MKANet in practical applications.

Data has now become a shortcoming of deep learning. Researchers in their own fields share the thinking that "deep neural networks might not always perform better when they eat more data," which still lacks experimental validation and a convincing guiding theory. Here to fill this lack, we design experiments from Identically Independent Distribution(IID) and Out of Distribution(OOD), which give powerful answers. For the purpose of guidance, based on the discussion of results, two theories are proposed: under IID condition, the amount of information determines the effectivity of each sample, the contribution of samples and difference between classes determine the amount of sample information and the amount of class information; under OOD condition, the cross-domain degree of samples determine the contributions, and the bias-fitting caused by irrelevant elements is a significant factor of cross-domain. The above theories provide guidance from the perspective of data, which can promote a wide range of practical applications of artificial intelligence.

Josephson junctions (JJs) with Josephson energy $E_J \lesssim 1K$ are widely employed as non-linear elements in superconducting circuits for quantum computing, operating at milli-Kelvin temperatures. Here we experimentally study incoherent phase slips (IPS) in low-$E_J$ Aluminum-based JJs at $T<0.2K$, where the IPS become the dominant source of dissipation. We observed strong suppression of the critical (switching) current and a very rapid growth of the zero-bias resistance with decreasing Josephson energy below $E_J \sim 1K$. This behavior is attributed to the IPSs whose rate exponentially increases with decreasing the ratio $E_J/T$. Our observations are in line with other data reported in literature. With further improvement of coherence of superconducting qubits, the observed dissipation from IPS might limit the performance of qubits based on low-$E_J$ junctions., Comment: 10 pages, 6 figures

An ultrafast spin current can be induced by femtosecond laser excitation in a ferromagnetic (FM) thin film in contact with a nonmagnetic (NM) metal. The propagation of an ultrafast spin current into NM metal has recently been found in experiments to generate transient spin accumulation. Unlike spin accumulation in equilibrium NM metals that occurs due to spin transport at the Fermi energy, transient spin accumulation involves highly nonequilibrium hot electrons well above the Fermi level. To date, the diffusion and dissipation of this transient spin accumulation has not been well studied. Using the superdiffusive spin transport model, we demonstrate how spin accumulation is generated in NM metals after laser excitation in an FM|NM bilayer. The spin accumulation shows an exponential decay from the FM|NM interface, with the decay length increasing to the maximum value and then decreasing until saturation. By analyzing the ultrafast dynamics of laser-excited hot electrons, the "effective mean free path", which can be characterized by the averaged product of the group velocity and lifetime of hot electrons, is found to play a key role. The interface reflectivity has little influence on the spin accumulation in NM metals. Our calculated results are in qualitative agreement with recent experiments., Comment: 9 pages, 5 figures, 9 equations

As a non-linear extension of the classic Linear Discriminant Analysis(LDA), Deep Linear Discriminant Analysis(DLDA) replaces the original Categorical Cross Entropy(CCE) loss function with eigenvalue-based loss function to make a deep neural network(DNN) able to learn linearly separable hidden representations. In this paper, we first point out DLDA focuses on training the cooperative discriminative ability of all the dimensions in the latent subspace, while put less emphasis on training the separable capacity of single dimension. To improve DLDA, a regularization method on within-class scatter matrix is proposed to strengthen the discriminative ability of each dimension, and also keep them complement each other. Experiment results on STL-10, CIFAR-10 and Pediatric Pneumonic Chest X-ray Dataset showed that our proposed regularization method Regularized Deep Linear Discriminant Analysis(RDLDA) outperformed DLDA and conventional neural network with CCE as objective. To further improve the discriminative ability of RDLDA in the local space, an algorithm named Subclass RDLDA is also proposed.

Given a sequence of positive Hermitian holomorphic line bundles $(L_p,h_p)$ on a K\"ahler manifold $X$, we establish the asymptotic expansion of the Bergman kernel of the space of global holomorphic sections of $L_p$, under a natural convergence assumption on the sequence of curvatures $c_1(L_p,h_p)$. We then apply this to study the asymptotic distribution of common zeros of random sequences of $m$-tuples of sections of $L_p$ as $p\to\infty$., Comment: 24 pages

Benefiting from the strong capabilities of deep CNNs for feature representation and nonlinear mapping, deep-learning-based methods have achieved excellent performance in single image super-resolution. However, most existing SR methods depend on the high capacity of networks which is initially designed for visual recognition, and rarely consider the initial intention of super-resolution for detail fidelity. Aiming at pursuing this intention, there are two challenging issues to be solved: (1) learning appropriate operators which is adaptive to the diverse characteristics of smoothes and details; (2) improving the ability of model to preserve the low-frequency smoothes and reconstruct the high-frequency details. To solve them, we propose a purposeful and interpretable detail-fidelity attention network to progressively process these smoothes and details in divide-and-conquer manner, which is a novel and specific prospect of image super-resolution for the purpose on improving the detail fidelity, instead of blindly designing or employing the deep CNNs architectures for merely feature representation in local receptive fields. Particularly, we propose a Hessian filtering for interpretable feature representation which is high-profile for detail inference, a dilated encoder-decoder and a distribution alignment cell to improve the inferred Hessian features in morphological manner and statistical manner respectively. Extensive experiments demonstrate that the proposed methods achieve superior performances over the state-of-the-art methods quantitatively and qualitatively. Code is available at https://github.com/YuanfeiHuang/DeFiAN., Comment: 14 pages, submitted to IEEE Transactions, codes are available at https://github.com/YuanfeiHuang/DeFiAN

This paper reviews the AIM 2020 challenge on efficient single image super-resolution with focus on the proposed solutions and results. The challenge task was to super-resolve an input image with a magnification factor x4 based on a set of prior examples of low and corresponding high resolution images. The goal is to devise a network that reduces one or several aspects such as runtime, parameter count, FLOPs, activations, and memory consumption while at least maintaining PSNR of MSRResNet. The track had 150 registered participants, and 25 teams submitted the final results. They gauge the state-of-the-art in efficient single image super-resolution.

This paper reviews the NTIRE 2020 challenge on perceptual extreme super-resolution with focus on proposed solutions and results. The challenge task was to super-resolve an input image with a magnification factor 16 based on a set of prior examples of low and corresponding high resolution images. The goal is to obtain a network design capable to produce high resolution results with the best perceptual quality and similar to the ground truth. The track had 280 registered participants, and 19 teams submitted the final results. They gauge the state-of-the-art in single image super-resolution., Comment: CVPRW 2020

Machine learning (ML) methods have been widely used in genomic studies. However, genomic data are often held by different stakeholders (e.g. hospitals, universities, and healthcare companies) who consider the data as sensitive information, even though they desire to collaborate. To address this issue, recent works have proposed solutions using Secure Multi-party Computation (MPC), which train on the decentralized data in a way that the participants could learn nothing from each other beyond the final trained model. We design and implement several MPC-friendly ML primitives, including class weight adjustment and parallelizable approximation of activation function. In addition, we develop the solution as an extension to TF Encrypted~\citep{dahl2018private}, enabling us to quickly experiment with enhancements of both machine learning techniques and cryptographic protocols while leveraging the advantages of TensorFlow's optimizations. Our implementation compares favorably with state-of-the-art methods, winning first place in Track IV of the iDASH2019 secure genome analysis competition., Comment: Description of the winning solution at Track IV of iDASH competition 2019, to be presented at the Trustworthy ML workshop co-located with ICLR2020

The spin- and energy-dependent interface reflectivity of a ferromagnetic (FM) film in contact with a nonmagnetic (NM) film is calculated using a first-principles transport method and incorporated into the superdiffusive spin transport model to study the femtosecond laser-induced ultrafast demagnetization of Fe|NM and Ni|NM (NM= Au, Al & Pt) bilayers. By comparing the calculated demagnetization with transparent and real interfaces, we demonstrate that the spin-dependent reflection of hot electrons has a noticeable influence on the ultrafast demagnetization and the associated terahertz electromagnetic radiation. In particular, a spin filtering effect is found at the Fe|NM interface that increases the spin current injected into the NM metal, which enhances both the resulting demagnetization and the resulting THz emission. This suggests that the THz radiation can be optimized by tailoring the interface, indicating a very large tunability., Comment: 14 pages, 16 figures

This paper reviews the AIM 2019 challenge on real world super-resolution. It focuses on the participating methods and final results. The challenge addresses the real world setting, where paired true high and low-resolution images are unavailable. For training, only one set of source input images is therefore provided in the challenge. In Track 1: Source Domain the aim is to super-resolve such images while preserving the low level image characteristics of the source input domain. In Track 2: Target Domain a set of high-quality images is also provided for training, that defines the output domain and desired quality of the super-resolved images. To allow for quantitative evaluation, the source input images in both tracks are constructed using artificial, but realistic, image degradations. The challenge is the first of its kind, aiming to advance the state-of-the-art and provide a standard benchmark for this newly emerging task. In total 7 teams competed in the final testing phase, demonstrating new and innovative solutions to the problem.

We have developed and characterized a symmetry-protected superconducting qubit that offers simultaneous exponential suppression of energy decay from charge and flux noise, and dephasing from flux noise. The qubit consists of a Cooper-pair box (CPB) shunted by a superinductor, thus forming a superconducting loop. Provided the offset charge on the CPB island is an odd number of electrons, the qubit potential corresponds to that of a $\cos \phi / 2$ Josephson element, preserving the parity of fluxons in the loop via Aharonov-Casher interference. In this regime, the logical-state wavefunctions reside in disjoint regions of phase space, thereby ensuring the protection against energy decay. By switching the protection on, we observed a ten-fold increase of the decay time, reaching up to $100 \mu \mathrm{s}$. Though the qubit is sensitive to charge noise, the sensitivity is much reduced in comparison with the charge qubit, and the charge-noise-induced dephasing time of the current device exceeds $1 \mu \mathrm{s}$. Implementation of the full dephasing protection can be achieved in the next-generation devices by combining several $\cos \phi / 2$ Josephson elements in a small array., Comment: 13 pages, 13 figures

We have designed superinductors made of strongly disordered superconductors for implementation in "hybrid" superconducting quantum circuits. The superinductors have been fabricated as meandered nanowires made of granular Aluminum films. Optimization of the device geometry enabled realization of superinductors with the inductance $\sim 1 {\mu}H$ and the self-resonance frequency over 3 GHz. These compact superinductors are attractive for a wide range of applications, from superconducting circuits for quantum computing to microwave elements of cryogenic parametric amplifiers and kinetic-inductance photon detectors., Comment: 5 pages, 4 figures; accepted manuscript, to be published in Phys. Rev. Appl

Recently, the deep convolutional neural network (CNN) has made remarkable progress in single image super resolution(SISR). However, blindly using the residual structure and dense structure to extract features from LR images, can cause the network to be bloated and difficult to train. To address these problems, we propose a simple and efficient distilling with residual network(DRN) for SISR. In detail, we propose residual distilling block(RDB) containing two branches, while one branch performs a residual operation and the other branch distills effective information. To further improve efficiency, we design residual distilling group(RDG) by stacking some RDBs and one long skip connection, which can effectively extract local features and fuse them with global features. These efficient features beneficially contribute to image reconstruction. Experiments on benchmark datasets demonstrate that our DRN is superior to the state-of-the-art methods, specifically has a better trade-off between performance and model size., Comment: 6 pages; Accepted to ICME2019

The accurate measurement of plasma potentials at high pressure is investigated with the emissive probe. The emissive probe I-V characteristic of argon plasma obtained at 100 Pa is interpreted in detail, showing two inflection points in the I-V trace. The accurate plasma potentials at high pressure can be determined from the first inflection point potentials (the higher potential) by the improved inflection point method of the emissive probe, while the second inflection point is a result of the additional ionizing phenomenon caused by the collision between the emitted electrons and the neutral argon atoms. Besides, the accuracy of the plasma potential measurement at high pressure is verified to be 0.3 V by the measurement of space potential distribution between two parallel plates in neutral gas at 100 Pa. The accurate measurement of space potentials in high pressure plasma and neutral gas indicates that the improved inflection point method of the emissive probe is promising in the study of collisional sheath structures and the electrostatic probe diagnostic of high pressure plasmas.

Learning fine-grained details is a key issue in image aesthetic assessment. Most of the previous methods extract the fine-grained details via random cropping strategy, which may undermine the integrity of semantic information. Extensive studies show that humans perceive fine-grained details with a mixture of foveal vision and peripheral vision. Fovea has the highest possible visual acuity and is responsible for seeing the details. The peripheral vision is used for perceiving the broad spatial scene and selecting the attended regions for the fovea. Inspired by these observations, we propose a Gated Peripheral-Foveal Convolutional Neural Network (GPF-CNN). It is a dedicated double-subnet neural network, i.e. a peripheral subnet and a foveal subnet. The former aims to mimic the functions of peripheral vision to encode the holistic information and provide the attended regions. The latter aims to extract fine-grained features on these key regions. Considering that the peripheral vision and foveal vision play different roles in processing different visual stimuli, we further employ a gated information fusion (GIF) network to weight their contributions. The weights are determined through the fully connected layers followed by a sigmoid function. We conduct comprehensive experiments on the standard AVA and Photo.net datasets for unified aesthetic prediction tasks: (i) aesthetic quality classification; (ii) aesthetic score regression; and (iii) aesthetic score distribution prediction. The experimental results demonstrate the effectiveness of the proposed method., Comment: Add more experiments

Annotating large scale datasets to train modern convolutional neural networks is prohibitively expensive and time-consuming for many real tasks. One alternative is to train the model on labeled synthetic datasets and apply it in the real scenes. However, this straightforward method often fails to generalize well mainly due to the domain bias between the synthetic and real datasets. Many unsupervised domain adaptation (UDA) methods are introduced to address this problem but most of them only focus on the simple classification task. In this paper, we present a novel UDA model to solve the more complex object detection problem in the context of autonomous driving. Our model integrates both pixel level and feature level based transformtions to fulfill the cross domain detection task and can be further trained end-to-end to pursue better performance. We employ objectives of the generative adversarial network and the cycle consistency loss for image translation in the pixel space. To address the potential semantic inconsistency problem, we propose region proposal based feature adversarial training to preserve the semantics of our target objects as well as further minimize the domain shifts. Extensive experiments are conducted on several different datasets, and the results demonstrate the robustness and superiority of our method.

We have studied superconducting coplanar-waveguide (CPW) resonators fabricated from disordered (granular) films of Aluminum. Very high kinetic inductance of these films, inherent to disordered materials, allows us to implement ultra-short (200 $\mu$m at a 5GHz resonance frequency) and high-impedance (up to 5 k$\Omega$) half-wavelength resonators. We have shown that the intrinsic losses in these resonators at temperatures $\lesssim 250$ mK are limited by resonator coupling to two-level systems in the environment. The demonstrated internal quality factors are comparable with those for CPW resonators made of conventional superconductors. High kinetic inductance and well-understood losses make these disordered Aluminum resonators promising for a wide range of microwave applications which include kinetic inductance photon detectors and superconducting quantum circuits.

We consider Einstein-Horndeski-Maxwell gravity, together with a cosmological constant and multiple Horndeski axions. We construct charged AdS planar black holes in general dimensions where the Horndeski anxions span over the planar directions. We analyse the thermodynamics and obtain the black hole volumes. We show that the reverse isoperimetric inequality can be violated, implying that these black holes can store information more efficiently than the Schwarzschild black hole., Comment: Latex, 25 pages, 1 figure, references added

We study the Berezin-Toeplitz quantization using as quantum space the space of eigenstates of the renormalized Bochner Laplacian corresponding to eigenvalues localized near the origin on a symplectic manifold. We show that this quantization has the correct semiclassical behavior and construct the corresponding star-product., Comment: 25 pages. arXiv admin note: text overlap with arXiv:0806.2370

We prove an estimate for Donaldson's $Q$-operator on a prequantized compact symplectic manifold. This estimate is an ingredient in the recent result of Keller and Lejmi about a symplectic generalization of Donaldson's lower bound for the $L^2$-norm of the Hermitian scalar curvature., Comment: 10 pages. To appear in Sci. China Math., 2017, Vol. 60

Existing object proposal algorithms usually search for possible object regions over multiple locations and scales separately, which ignore the interdependency among different objects and deviate from the human perception procedure. To incorporate global interdependency between objects into object localization, we propose an effective Tree-structured Reinforcement Learning (Tree-RL) approach to sequentially search for objects by fully exploiting both the current observation and historical search paths. The Tree-RL approach learns multiple searching policies through maximizing the long-term reward that reflects localization accuracies over all the objects. Starting with taking the entire image as a proposal, the Tree-RL approach allows the agent to sequentially discover multiple objects via a tree-structured traversing scheme. Allowing multiple near-optimal policies, Tree-RL offers more diversity in search paths and is able to find multiple objects with a single feed-forward pass. Therefore, Tree-RL can better cover different objects with various scales which is quite appealing in the context of object proposal. Experiments on PASCAL VOC 2007 and 2012 validate the effectiveness of the Tree-RL, which can achieve comparable recalls with current object proposal algorithms via much fewer candidate windows., Comment: Advances in Neural Information Processing Systems 2016

It is known that a compact symplectic manifold endowed with a prequantum line bundle can be embedded in the projective space generated by the eigensections of low energy of the Bochner Laplacian acting on high $p$-tensor powers of the prequantum line bundle. We show that the Fubini-Study forms induced by these embeddings converge at speed rate $1/p^{2}$ to the symplectic form. This result implies the generalization to the almost-K\"ahler case of the lower bounds on the Calabi functional given by Donaldson for K\"ahler manifolds, as shown by Lejmi and Keller., Comment: 23 pages; v.2 is a final update to agree with the published paper

The neutral component of an inert scalar multiplet with hypercharge can provide a stable dark matter particle when its real and imaginary parts have a splitting mass spectrum. Otherwise, a tree-level dark matter-nucleon scattering mediated by the $Z$ boson will be much above the experimental limit. In this paper we focus on a mixed inert scalar triplet dark matter scenario where a complex scalar triplet with hypercharge can mix with another real scalar triplet without hypercharge through their renormalizable coupling to the standard model Higgs doublet. We consider three specified cases that carry most of the relevant features of the full parameter space: (i) the neutral component of the real triplet dominates the dark matter particle, (ii) the neutral component of the complex triplet dominates the dark matter particle; and (iii) the neutral components of the real and complex triplets equally constitute the dark matter particle. Subject to the dark matter relic abundance and direct detection constraint, we perform a systematic study on the allowed parameter space with particular emphasis on the interplay among triplet-doublet terms and gauge interactions. In the presence of these mixed inert scalar triplets, some heavy Dirac fermions composed of inert fermion doublets can be utilized to generate a tiny Majorana neutrino mass term at one-loop level and realize a successful leptogenesis for explaining the cosmic baryon asymmetry., Comment: 38 pages, 23 figures, typos are corrected and references are added

Spurred by the dramatic mobile IP growth and the emerging Internet of Things (IoT) and cloud-based applications, wireless networking is witnessing a paradigm shift. By fully exploiting the spatial degrees of freedom, the massive multipleinput- multiple-output (MIMO) technology promises significant gains in both data rates and link reliability. This paper presents a time-division duplex (TDD)-based 128-antenna massive MIMO prototyping system designed to operate on a 20 MHz bandwidth. Up to twelve single-antenna users can be served by the designed system at the same time. System model is provided and link-level simulation corresponding to our practical TDDbased massive MIMO prototyping system is conducted to validate our design and performance of the algorithms. Based on the system hardware design demonstrated in this paper, both uplink real-time video and downlink data transmissions are realized, and the experiment results show that 268.8 Mbps rate was achieved for eight single-antenna users using QPSK modulation. The maximum spectral efficiency of the designed system will be 80.64 bit/s/Hz by twelve single-antenna users with 256-QAM modulation.

We extend the standard model by three types of inert fields including Majorana fermion singlets/triplets, real Higgs singlets/triplets and leptonic Higgs doublets. In the presence of a softly broken lepton number and an exactly conserved Z_2 discrete symmetry, these inert fields together can mediate a one-loop diagram for a Majorana neutrino mass generation. The heavier inert fields can decay to realize a successful leptogenesis while the lightest inert field can provide a stable dark matter candidate. As an example, we demonstrate the leptogenesis by the inert Higgs doublet decays. We also perform a systematic study on the inert Higgs triplet dark matter scenario where the interference between the gauge and Higgs portal interactions can significantly affect the dark matter properties., Comment: 12 pages, 7 figures, references added

Object proposal is essential for current state-of-the-art object detection pipelines. However, the existing proposal methods generally fail in producing results with satisfying localization accuracy. The case is even worse for small objects which however are quite common in practice. In this paper we propose a novel Scale-aware Pixel-wise Object Proposal (SPOP) network to tackle the challenges. The SPOP network can generate proposals with high recall rate and average best overlap (ABO), even for small objects. In particular, in order to improve the localization accuracy, a fully convolutional network is employed which predicts locations of object proposals for each pixel. The produced ensemble of pixel-wise object proposals enhances the chance of hitting the object significantly without incurring heavy extra computational cost. To solve the challenge of localizing objects at small scale, two localization networks which are specialized for localizing objects with different scales are introduced, following the divide-and-conquer philosophy. Location outputs of these two networks are then adaptively combined to generate the final proposals by a large-/small-size weighting network. Extensive evaluations on PASCAL VOC 2007 show the SPOP network is superior over the state-of-the-art models. The high-quality proposals from SPOP network also significantly improve the mean average precision (mAP) of object detection with Fast-RCNN framework. Finally, the SPOP network (trained on PASCAL VOC) shows great generalization performance when testing it on ILSVRC 2013 validation set., Comment: accepted by IEEE Transactions on Image Processing

In this paper, we deal with a class of mean-field backward stochastic differential equations (BSDEs) related to finite state, continuous time Markov chains. We obtain the existence and uniqueness theorem and a comparison theorem for solutions of one-dimensional mean-field BSDEs under Lipschitz condition.

In this paper, we deal with a class of mean-field backward stochastic differential equations with subdifferrential operator corresponding to a lower semi-continuous convex function. By means of Yosida approximation, the existence and uniqueness of the solution is established. As an application, we give a probability interpretation for the viscosity solutions of a class of nonlocal parabolic variational inequalities.

In this paper, we deal with a class of multivalued backward doubly stochastic differential equations with time delayed coefficients. Based on a slight extension of the existence and uniqueness of solutions for backward doubly stochastic differential equations with time delayed coefficients, we establish the existence and uniqueness of solutions for these equations by means of Yosida approximation., Comment: 12 pages

We calculate the second coefficient of the asymptotic expansion of the Bergman kernel of the Hodge-Dolbeault operator associated to high powers of a Hermitian line bundle with non-degenerate curvature, using the method of formal power series developed by Ma and Marinescu., Comment: 32 pages

Given a smooth compact manifold with boundary, we show that the subcomplex of the deformed de Rham complex consisting of eigenspaces of small eigenvalues of the Witten Laplacian is canonically isomorphic to the Thom-Smale complex constructed by Laudenbach. Our proof is based on Bismut-Lebeau's analytic localization techniques. As a by-product, we obtain Morse inequalities for manifolds with boundary., Comment: 20 pages

In this paper, we prove equivariant Morse inequalities via Bismut-Lebeau's analytic localization techniques. As an application, we obtain Morse inequalities on compact manifold with nonempty boundary by applying equivariant Morse inequalities to the doubling manifold., Comment: 19 pages

In this paper, we study a class of backward stochastic Volterra integral equations driven by Teugels martingales associated with an independent L\'{e}vy process and an independent Brownian motion (BSVIELs). We prove the existence and uniqueness as well as stability of the adapted M-solutions for those equations. Moreover, a duality principle and then a comparison theorem are established. As an application, we derive a class of dynamic risk measures by means of M-solutions of certain BSVIELs.

Thermoelectric power of a material, typically governed by its band structure and carrier density, can be varied by chemical doping that is often restricted by solubility of the dopant. Materials showing large thermoelectric power are useful for many industrial applications, such as the heat-to-electricity conversion and the thermoelectric cooling device. Here we show a full electric field tuning of thermoelectric power in a dual-gated bilayer graphene device resulting from the opening of a band-gap by applying a perpendicular electric field on bilayer graphene. We uncover a large enhancement in thermoelectric power at low temperature, which may open up a new possibility in low temperature thermoelectric application using graphene-based device., Comment: 12 pages, 4 figures