Searcharxiv⌕ Search

arXiv subjects

Jie Hu

Publications and source records attributed to Jie Hu.

At least 163 records · Page 9Linked to original sources
arXiv2023

GenDet: Towards Good Generalizations for AI-Generated Image Detection

The misuse of AI imagery can have harmful societal effects, prompting the creation of detectors to combat issues like the spread of fake news. Existing methods can effectively detect images generated by seen generators, but it is challenging to detect those generated by unseen generators. They do not concentrate on amplifying the output discrepancy when detectors process real versus fake images. This results in a close output distribution of real and fake samples, increasing classification difficulty in detecting unseen generators. This paper addresses the unseen-generator detection problem by considering this task from the perspective of anomaly detection and proposes an adversarial teacher-student discrepancy-aware framework. Our method encourages smaller output discrepancies between the student and the teacher models for real images while aiming for larger discrepancies for fake images. We employ adversarial learning to train a feature augmenter, which promotes smaller discrepancies between teacher and student networks when the inputs are fake images. Our method has achieved state-of-the-art on public benchmarks, and the visualization results show that a large output discrepancy is maintained when faced with various types of generators.

cs.CV↗
arXiv2023

Development of MKIDs in the Optical and Near-infrared Bands for SPIAKID

SpectroPhotometric Imaging in Astronomy with Kinetic Inductance Detectors (SPIAKID) aims at designing, building, and deploying on the sky a spectrophotometric imager based on microwave kinetic inductance detectors (MKIDs) in the optical and near-infrared bands. MKIDs show a fast response and the ability to resolve photon energy compared to the conventional Charge-coupled Devices (CCDs). In this paper, we present the design and simulation of the MKID arrays for SPIAKID. The detectors consist of four arrays with each array of 20,000 lumped-element pixels, and each array will be read with 10 readout lines. %The array is designed to have resonances between 4-8GHz with a frequency spacing of 2 MHz and a coupling quality factor (Qc) of about 50000. The meander material of the resonators is trilayer TiN/Ti/TiN to have better uniformity of the critical temperature across the array. We also present the measurement result for a test array with $30\times30$ pixels which is a subset of the designed 2000-pixel array to verify the design and fabrication. The current measured best energy resolving power $R = E/ΔE$ is 2.4 at $λ= 405~$nm and the current medium R is around 1.7. We have also observed the response of the TiN/Ti/TiN is much smaller than expected.

astro-ph.IM↗
arXiv2023

Massive Wireless Energy Transfer without Channel State Information via Imperfect Intelligent Reflecting Surfaces

Intelligent Reflecting Surface (IRS) utilizes low-cost, passive reflecting elements to enhance the passive beam gain, improve Wireless Energy Transfer (WET) efficiency, and enable its deployment for numerous Internet of Things (IoT) devices. However, the increasing number of IRS elements presents considerable channel estimation challenges. This is due to the lack of active Radio Frequency (RF) chains in an IRS, while pilot overhead becomes intolerable. To address this issue, we propose a Channel State Information (CSI)-free scheme that maximizes received energy in a specific direction and covers the entire space through phased beam rotation. Furthermore, we take into account the impact of an imperfect IRS and meticulously design the active precoder and IRS reflecting phase shift to mitigate its effects. Our proposed technique does not alter the existing IRS hardware architecture, allowing for easy implementation in the current system, and enabling access or removal of any Energy Receivers (ERs) without additional cost. Numerical results illustrate the efficacy of our CSI-free scheme in facilitating large-scale IRS without compromising performance due to excessive pilot overhead. Furthermore, our scheme outperforms the CSI-based counterpart in scenarios involving large-scale ERs, making it a promising solution in the era of IoT.

eess.SP↗
arXiv2023

Learning-Based Biharmonic Augmentation for Point Cloud Classification

Point cloud datasets often suffer from inadequate sample sizes in comparison to image datasets, making data augmentation challenging. While traditional methods, like rigid transformations and scaling, have limited potential in increasing dataset diversity due to their constraints on altering individual sample shapes, we introduce the Biharmonic Augmentation (BA) method. BA is a novel and efficient data augmentation technique that diversifies point cloud data by imposing smooth non-rigid deformations on existing 3D structures. This approach calculates biharmonic coordinates for the deformation function and learns diverse deformation prototypes. Utilizing a CoefNet, our method predicts coefficients to amalgamate these prototypes, ensuring comprehensive deformation. Moreover, we present AdvTune, an advanced online augmentation system that integrates adversarial training. This system synergistically refines the CoefNet and the classification network, facilitating the automated creation of adaptive shape deformations contingent on the learner status. Comprehensive experimental analysis validates the superiority of Biharmonic Augmentation, showcasing notable performance improvements over prevailing point cloud augmentation techniques across varied network designs.

cs.CV↗
arXiv2023

A robust version of the multipartite Hajnal--Szemerédi theorem

In this note we show the following strengthening of a multipartite version of the Hajnal--Szemerédi theorem. For an integer $r \ge 3$ and $γ> 0$, there exists a constant $C$ such that if $p\ge Cn^{-2/r}(\log n)^{1/{r \choose 2}}$ and $G$ is a balanced $r$-partite graph with each vertex class of size $n$ and $δ^\ast(G)\ge (1-1/r+γ)n$, then with high probability the random subgraph $G(p)$ of $G$ contains a $K_r$-factor. We also use it to derive corresponding transversal versions.

math.CO↗
arXiv2023

Multi-Domain Polarization for Enhancing the Physical Layer Security of MIMO Systems

A novel Physical Layer Security (PLS) framework is conceived for enhancing the security of the wireless communication systems by exploiting multi-domain polarization in Multiple-Input Multiple-Output (MIMO) systems. We design a sophisticated key generation scheme based on multi-domain polarization, and the corresponding receivers. An in-depth analysis of the system's secrecy rate is provided, demonstrating the confidentiality of our approach in the presence of eavesdroppers having strong computational capabilities. More explicitly, our simulation results and theoretical analysis corroborate the advantages of the proposed scheme in terms of its bit error rate (BER), block error rate (BLER), and maximum achievable secrecy rate. Our findings indicate that the innovative PLS framework effectively enhances the security and reliability of wireless communication systems. For instance, in a $4\times4$ MIMO setup, the proposed PLS strategy exhibits an improvement of $2$dB compared to conventional MIMO, systems at a BLER of $2\cdot 10^{-5}$ while the eavesdropper's BLER reaches $1$.

cs.IT↗
arXiv2023

Spatio-Temporal Meta Contrastive Learning

Spatio-temporal prediction is crucial in numerous real-world applications, including traffic forecasting and crime prediction, which aim to improve public transportation and safety management. Many state-of-the-art models demonstrate the strong capability of spatio-temporal graph neural networks (STGNN) to capture complex spatio-temporal correlations. However, despite their effectiveness, existing approaches do not adequately address several key challenges. Data quality issues, such as data scarcity and sparsity, lead to data noise and a lack of supervised signals, which significantly limit the performance of STGNN. Although recent STGNN models with contrastive learning aim to address these challenges, most of them use pre-defined augmentation strategies that heavily depend on manual design and cannot be customized for different Spatio-Temporal Graph (STG) scenarios. To tackle these challenges, we propose a new spatio-temporal contrastive learning (CL4ST) framework to encode robust and generalizable STG representations via the STG augmentation paradigm. Specifically, we design the meta view generator to automatically construct node and edge augmentation views for each disentangled spatial and temporal graph in a data-driven manner. The meta view generator employs meta networks with parameterized generative model to customize the augmentations for each input. This personalizes the augmentation strategies for every STG and endows the learning framework with spatio-temporal-aware information. Additionally, we integrate a unified spatio-temporal graph attention network with the proposed meta view generator and two-branch graph contrastive learning paradigms. Extensive experiments demonstrate that our CL4ST significantly improves performance over various state-of-the-art baselines in traffic and crime prediction.

cs.LG↗
arXiv2023

Robust NOMA-assisted OTFS-ISAC Network Design with 3D Motion Prediction Topology

This paper proposes a novel non-orthogonal multiple access (NOMA)-assisted orthogonal time-frequency space (OTFS)-integrated sensing and communication (ISAC) network, which uses unmanned aerial vehicles (UAVs) as air base stations to support multiple users. By employing ISAC, the UAV extracts position and velocity information from the user's echo signals, and non-orthogonal power allocation is conducted to achieve a superior achievable rate. A 3D motion prediction topology is used to guide the NOMA transmission for multiple users, and a robust power allocation solution is proposed under perfect and imperfect channel estimation for Maxi-min Fairness (MMF) and Maximum sum-Rate (SR) problems. Simulation results demonstrate the superiority of the proposed NOMA-assisted OTFS-ISAC system over other systems in terms of achievable rate under both perfect and imperfect channel conditions with the aid of 3D motion prediction topology.

eess.SP↗
arXiv2023

Green Beamforming Design for Integrated Sensing and Communication Systems: A Practical Approach Using Beam-Matching Error Metrics

In this paper, we propose a green beamforming design for the integrated sensing and communication (ISAC) system, using beam-matching error to assess radar performance. The beam-matching error metric, which considers the mean square error between the desired and designed beam patterns, provides a more practical evaluation approach. To tackle the non-convex challenge inherent in beamforming design, we apply semidefinite relaxation (SDR) to address the rank-one relaxation issue, followed by the iterative rank minimization algorithm (IRM) for rank-one recovery. The simulation results showcase the effectiveness of our proposed optimal beamforming design, emphasizing the exceptional performance of the radar component in sensing tasks.

eess.SP↗
arXiv2023

Reconfigurable Intelligent Sensing Surface aided Wireless Powered Communication Networks: A Sensing-Then-Reflecting Approach

This paper presents a reconfigurable intelligent sensing surface (RISS) that combines passive and active elements to achieve simultaneous reflection and direction of arrival (DOA) estimation tasks. By utilizing DOA information from the RISS instead of conventional channel estimation, the pilot overhead is reduced and the RISS becomes independent of the hybrid access point (HAP), enabling efficient operation. Specifically, the RISS autonomously estimates the DOA of uplink signals from single-antenna users and reflects them using the HAP's slowly varying DOA information. During downlink transmission, it updates the HAP's DOA information and designs the reflection phase of energy signals based on the latest user DOA information. The paper includes a comprehensive performance analysis, covering system design, protocol details, receiving performance, and RISS deployment suggestions. We derive a closed-form expression to analyze system performance under DOA errors, and calculate the statistical distribution of user received energy using the moment-matching technique. We provide a recommended transmit power to meet a specified outage probability and energy threshold. Numerical results demonstrate that the proposed system outperforms the conventional counterpart by 2.3 dB and 4.7 dB for Rician factors $κ_h=κ_G=1$ and $κ_h=κ_G=10$, respectively.

cs.IT↗
arXiv2023

IFT: Image Fusion Transformer for Ghost-free High Dynamic Range Imaging

Multi-frame high dynamic range (HDR) imaging aims to reconstruct ghost-free images with photo-realistic details from content-complementary but spatially misaligned low dynamic range (LDR) images. Existing HDR algorithms are prone to producing ghosting artifacts as their methods fail to capture long-range dependencies between LDR frames with large motion in dynamic scenes. To address this issue, we propose a novel image fusion transformer, referred to as IFT, which presents a fast global patch searching (FGPS) module followed by a self-cross fusion module (SCF) for ghost-free HDR imaging. The FGPS searches the patches from supporting frames that have the closest dependency to each patch of the reference frame for long-range dependency modeling, while the SCF conducts intra-frame and inter-frame feature fusion on the patches obtained by the FGPS with linear complexity to input resolution. By matching similar patches between frames, objects with large motion ranges in dynamic scenes can be aligned, which can effectively alleviate the generation of artifacts. In addition, the proposed FGPS and SCF can be integrated into various deep HDR methods as efficient plug-in modules. Extensive experiments on multiple benchmarks show that our method achieves state-of-the-art performance both quantitatively and qualitatively.

cs.CV↗
arXiv2023

An Edge-based Graph Transformer Network for Anti-Cancer Drug Synergy Prediction

Drug combination therapy is a powerful solution for the treatment of complex disease such as cancers due to its capability of therapeutic efficacy and reducing side effects. Nevertheless, it is very difficult to screen all drug combinations by experiments since the vast number of possible combinations. Currently, computational methods, especially graph neural networks and transformer, have been developed to discover the prioritization of drug combinations and shown promising potentials. Despite great achievements have been obtained by existing computational models, they all neglected high-order semantic information of drugs and the importance of the chemical bond features, which contained rich information and is represented by edge of graphs in drug predictions. In this work, we present a novel model named EGTSyn (Edge-based Graph Transformer network for drug Synergy prediction) for anti-cancer drug synergistic effect prediction. We design an EGNN (edge-based graph neural network) module and a GTDblock (Graph Transformer for Drugs block). EGNN is employed to capture the global structure information of the chemicals as well as the importance of chemical bonds that has been neglected by most of the previous studies. GTDblock combines the EGNN module with a Transformer-architecture encoder to extract high-order semantic information of drugs.

cs.CE↗
arXiv2023

Information Theory-Guided Heuristic Progressive Multi-View Coding

Multi-view representation learning captures comprehensive information from multiple views of a shared context. Recent works intuitively apply contrastive learning (CL) to learn representations, regarded as a pairwise manner, which is still scalable: view-specific noise is not filtered in learning view-shared representations; the fake negative pairs, where the negative terms are actually within the same class as the positive, and the real negative pairs are coequally treated; and evenly measuring the similarities between terms might interfere with optimization. Importantly, few works research the theoretical framework of generalized self-supervised multi-view learning, especially for more than two views. To this end, we rethink the existing multi-view learning paradigm from the information theoretical perspective and then propose a novel information theoretical framework for generalized multi-view learning. Guided by it, we build a multi-view coding method with a three-tier progressive architecture, namely Information theory-guided heuristic Progressive Multi-view Coding (IPMC). In the distribution-tier, IPMC aligns the distribution between views to reduce view-specific noise. In the set-tier, IPMC builds self-adjusted pools for contrasting, which utilizes a view filter to adaptively modify the pools. Lastly, in the instance-tier, we adopt a designed unified loss to learn discriminative representations and reduce the gradient interference. Theoretically and empirically, we demonstrate the superiority of IPMC over state-of-the-art methods.

cs.CV↗
arXiv2023

Bayesian Inference of Phenotypic Plasticity of Cancer Cells Based on Dynamic Model for Temporal Cell Proportion Data

Mounting evidence underscores the prevalent hierarchical organization of cancer tissues. At the foundation of this hierarchy reside cancer stem cells, a subset of cells endowed with the pivotal role of engendering the entire cancer tissue through cell differentiation. In recent times, substantial attention has been directed towards the phenomenon of cancer cell plasticity, where the dynamic interconversion between cancer stem cells and non-stem cancer cells has garnered significant interest. Since the task of detecting cancer cell plasticity from empirical data remains a formidable challenge, we propose a Bayesian statistical framework designed to infer phenotypic plasticity within cancer cells, utilizing temporal data on cancer stem cell proportions. Our approach is grounded in a stochastic model, adept at capturing the dynamic behaviors of cells. Leveraging Bayesian analysis, we explore the moment equation governing cancer stem cell proportions, derived from the Kolmogorov forward equation of our stochastic model. With improved Euler method for ordinary differential equations, a new statistical method for parameter estimation in nonlinear ordinary differential equations models is developed, which also provides novel ideas for the study of compositional data. Extensive simulations robustly validate the efficacy of our proposed method. To further corroborate our findings, we apply our approach to analyze published data from SW620 colon cancer cell lines. Our results harmonize with \emph{in situ} experiments, thereby reinforcing the utility of our method in discerning and quantifying phenotypic plasticity within cancer cells.

stat.AP↗
arXiv2023

Pseudo-label Alignment for Semi-supervised Instance Segmentation

Pseudo-labeling is significant for semi-supervised instance segmentation, which generates instance masks and classes from unannotated images for subsequent training. However, in existing pipelines, pseudo-labels that contain valuable information may be directly filtered out due to mismatches in class and mask quality. To address this issue, we propose a novel framework, called pseudo-label aligning instance segmentation (PAIS), in this paper. In PAIS, we devise a dynamic aligning loss (DALoss) that adjusts the weights of semi-supervised loss terms with varying class and mask score pairs. Through extensive experiments conducted on the COCO and Cityscapes datasets, we demonstrate that PAIS is a promising framework for semi-supervised instance segmentation, particularly in cases where labeled data is severely limited. Notably, with just 1\% labeled data, PAIS achieves 21.2 mAP (based on Mask-RCNN) and 19.9 mAP (based on K-Net) on the COCO dataset, outperforming the current state-of-the-art model, \ie, NoisyBoundary with 7.7 mAP, by a margin of over 12 points. Code is available at: \url{https://github.com/hujiecpp/PAIS}.

cs.CV↗
arXiv2023

Single-sensor and real-time ultrasonic imaging using an AI-driven disordered metasurface

Non-destructive testing and medical diagnostic techniques using ultrasound has become indispensable in evaluating the state of materials or imaging the internal human body, respectively. To conduct spatially resolved high-quality observations, conventionally, sophisticated phased arrays are used both at the emitting and receiving ends of the setup. In comparison, single-sensor imaging techniques offer significant benefits including compact physical dimensions and reduced manufacturing expenses. However, recent advances such as compressive sensing have shown that this improvement comes at the cost of additional time-consuming dynamic spatial scanning or multi-mode mask switching, which severely hinders the quest for real-time imaging. Consequently, real-time single-sensor imaging, at low cost and simple design, still represents a demanding and largely unresolved challenge till this day. Here, we bestow on ultrasonic metasurface with both disorder and artificial intelligence (AI). The former ensures strong dispersion and highly complex scattering to encode the spatial information into frequency spectra at an arbitrary location, while the latter is used to decode instantaneously the amplitude and spectral component of the sample under investigation. Thus, thanks to this symbiosis, we demonstrate that a single fixed sensor suffices to recognize complex ultrasonic objects through the random scattered field from an unpretentious metasurface, which enables real-time and low-cost imaging, easily extendable to 3D.

physics.app-ph↗
arXiv2023

GenImage: A Million-Scale Benchmark for Detecting AI-Generated Image

The extraordinary ability of generative models to generate photographic images has intensified concerns about the spread of disinformation, thereby leading to the demand for detectors capable of distinguishing between AI-generated fake images and real images. However, the lack of large datasets containing images from the most advanced image generators poses an obstacle to the development of such detectors. In this paper, we introduce the GenImage dataset, which has the following advantages: 1) Plenty of Images, including over one million pairs of AI-generated fake images and collected real images. 2) Rich Image Content, encompassing a broad range of image classes. 3) State-of-the-art Generators, synthesizing images with advanced diffusion models and GANs. The aforementioned advantages allow the detectors trained on GenImage to undergo a thorough evaluation and demonstrate strong applicability to diverse images. We conduct a comprehensive analysis of the dataset and propose two tasks for evaluating the detection method in resembling real-world scenarios. The cross-generator image classification task measures the performance of a detector trained on one generator when tested on the others. The degraded image classification task assesses the capability of the detectors in handling degraded images such as low-resolution, blurred, and compressed images. With the GenImage dataset, researchers can effectively expedite the development and evaluation of superior AI-generated image detectors in comparison to prevailing methodologies.

cs.CV↗
arXiv2023

Applied Regression Analysis of Correlations for Correlated Data

Correlated data are ubiquitous in today's data-driven society. While regression models for analyzing means and variances of responses of interest are relatively well-developed, the development of these models for analyzing the correlations is largely confined to longitudinal data, a special form of sequentially correlated data. This paper proposes a new method for the analysis of correlations to fully exploit the use of covariates for general correlated data. In a renewed analysis of the Classroom data, a highly unbalanced multilevel clustered data with within-class and within-school correlations, our method reveals informative insights on these structures not previously known. In another analysis of the malaria immune response data in Benin, a longitudinal study with time-dependent covariates where the exact times of the observations are not available, our approach again provides promising new results. At the heart of our approach is a new generalized z-transformation that converts correlation matrices constrained to be positive definite to vectors with unrestricted support, and is order-invariant. These two properties enable us to develop regression analysis incorporating covariates for the modelling of correlations via the use of maximum likelihood.

stat.ME↗