Search

Recent progress in image deblurring techniques focuses mainly on operating in both frequency and spatial domains using the Fourier transform (FT) properties. However, their performance is limited due to the dependency of FT on stationary signals and its lack of capability to extract spatial-frequency properties. In this paper, we propose a novel approach based on the Fractional Fourier Transform (FRFT), a unified spatial-frequency representation leveraging both spatial and frequency components simultaneously, making it ideal for processing non-stationary signals like images. Specifically, we introduce a Fractional Fourier Transformer (F2former), where we combine the classical fractional Fourier based Wiener deconvolution (F2WD) as well as a multi-branch encoder-decoder transformer based on a new fractional frequency aware transformer block (F2TB). We design F2TB consisting of a fractional frequency aware self-attention (F2SA) to estimate element-wise product attention based on important frequency components and a novel feed-forward network based on frequency division multiplexing (FM-FFN) to refine high and low frequency features separately for efficient latent clear image restoration. Experimental results for the cases of both motion deblurring as well as defocus deblurring show that the performance of our proposed method is superior to other state-of-the-art (SOTA) approaches., Comment: 20 pages, 21 figures

Self-supervised learning (SSL) has emerged as a promising paradigm for medical image analysis by harnessing unannotated data. Despite their potential, the existing SSL approaches overlook the high anatomical similarity inherent in medical images. This makes it challenging for SSL methods to capture diverse semantic content in medical images consistently. This work introduces a novel and generalized solution that implicitly exploits anatomical similarities by integrating codebooks in SSL. The codebook serves as a concise and informative dictionary of visual patterns, which not only aids in capturing nuanced anatomical details but also facilitates the creation of robust and generalized feature representations. In this context, we propose CoBooM, a novel framework for self-supervised medical image learning by integrating continuous and discrete representations. The continuous component ensures the preservation of fine-grained details, while the discrete aspect facilitates coarse-grained feature extraction through the structured embedding space. To understand the effectiveness of CoBooM, we conduct a comprehensive evaluation of various medical datasets encompassing chest X-rays and fundus images. The experimental results reveal a significant performance gain in classification and segmentation tasks., Comment: Accepted in MICCAI 2024

In this study, we present a novel approach for predicting genomic information from medical imaging modalities using a transformer-based model. We aim to bridge the gap between imaging and genomics data by leveraging transformer networks, allowing for accurate genomic profile predictions from CT/MRI images. Presently most studies rely on the use of whole slide images (WSI) for the association, which are obtained via invasive methodologies. We propose using only available CT/MRI images to predict genomic sequences. Our transformer based approach is able to efficiently generate associations between multiple sequences based on CT/MRI images alone. This work paves the way for the use of non-invasive imaging modalities for precise and personalized healthcare, allowing for a better understanding of diseases and treatment.

As our cities become more intelligent and more connected with new technologies like 6G, improving communication between vehicles and infrastructure is essential while reducing energy consumption. This study proposes a secure framework for vehicle-to-infrastructure (V2I) backscattering near an eavesdropping vehicle to maximize the sum secrecy rate of V2I backscatter communication over multiple coherence slots. This sustainable framework aims to jointly optimize the reflection coefficients at the backscattering vehicle, carrier emitter power, and artificial noise at the infrastructure, along with the target vehicle's linear trajectory in the presence of an eavesdropping vehicle in the parallel lane. To achieve this optimization, we separated the problem into three parts: backscattering coefficient, power allocation, and trajectory design problems. We respectively adopted parallel computing, fractional programming, and finding all the candidates for the global optimal solution to obtain the global optimal solution for these three problems. Our simulations verified the fast convergence of our alternating optimization algorithm and showed that our proposed secure V2I backscattering outperforms the existing benchmark by over 4.7 times in terms of secrecy rate for 50 slots. Overall, this fundamental research on V2I backscattering provided insights to improve vehicular communication's connectivity, efficiency, and security., Comment: This paper contains 6 pages and 6 figures, which are accepted and will be presented in the proceedings of the 2024 IEEE International Conference on Communications (ICC), but not online yet

The dynamics of the state-switching process of black holes in dRGT massive gravity theory is presented using free energy landscape and stochastic Langevin equations. The free energy landscape is constructed using the Gibbons-Hawking path integral method. The black hole phases are characterized by taking its horizon radius as the order parameter. The free energy landscape provides three black hole phases: small, intermediate, and large. The small and large black holes are thermodynamically stable whereas the intermediate one is unstable. The Martin-Siggia-Rose-Janssen-de Dominicis (MSRJD) functional describes the stochastic dynamics of black hole phase transition. The Hamiltonian flow lines are obtained from the MSRJD functional and are used to analyze the stability and the phase transition properties. The dominant kinetic path between different phases is discussed for various configurations of the free energy landscape. We discuss the effect of black hole charge and the graviton mass on the critical behavior of black hole phase transition., Comment: 13 pages, 10 figures

This paper studies backscatter communication (BackCom) systems with a passive backscatter tag. The effectiveness of these tags is limited by the amount of energy they can harness from incident radio signals, which are used to backscatter information through the modulation of reflections. To address this limitation, we adopt a practical Constant-Linear-Constant (CLC) energy harvesting model that accounts for the harvester's sensitivity and saturation threshold, both of which depend on the input power. This paper aims to maximize this harvested power at a passive tag by optimally designing the underlying M-ary amplitude-shift keying (ASK) modulator in a monostatic BackCom system. Specifically, we derive the closed-form expression for the global optimal reflection coefficients that maximize the tag's harvested power while satisfying the minimum symbol error rate (SER) requirement, tag sensitivity, and reader sensitivity constraints. We also proposed optimal binary-ASK modulation design to gain novel design insights on practical BackCom systems with readers having superior sensitivity. We have validated these nontrivial analytical claims via extensive simulations. The numerical results provide insight into the impact of the transmit symbol probability, tag sensitivity constraint, and SER on the maximum average harvested power. Remarkably, our design achieves an overall gain of around 13% over the benchmark, signifying its utility in improving the efficiency of BackCom systems. Moreover, our proposed solution methodology for determining the maximum average harvested power is applicable to any type of energy harvesting model that exhibits a monotonic increasing relationship with the input power., Comment: 16 pages, 15 figures, journal

Self-supervised learning (SSL) has emerged as a promising technique for medical image analysis due to its ability to learn without annotations. However, despite the promising potential, conventional SSL methods encounter limitations, including challenges in achieving semantic alignment and capturing subtle details. This leads to suboptimal representations, which fail to accurately capture the underlying anatomical structures and pathological details. In response to these constraints, we introduce a novel SSL framework OPTiML, employing optimal transport (OT), to capture the dense semantic invariance and fine-grained details, thereby enhancing the overall effectiveness of SSL in medical image representation learning. The core idea is to integrate OT with a cross-viewpoint semantics infusion module (CV-SIM), which effectively captures complex, fine-grained details inherent in medical images across different viewpoints. In addition to the CV-SIM module, OPTiML imposes the variance and covariance regularizations within OT framework to force the model focus on clinically relevant information while discarding less informative features. Through these, the proposed framework demonstrates its capacity to learn semantically rich representations that can be applied to various medical imaging tasks. To validate its effectiveness, we conduct experimental studies on three publicly available datasets from chest X-ray modality. Our empirical results reveal OPTiML's superiority over state-of-the-art methods across all evaluated tasks.

Self-supervised learning (SSL) is potentially useful in reducing the need for manual annotation and making deep learning models accessible for medical image analysis tasks. By leveraging the representations learned from unlabeled data, self-supervised models perform well on tasks that require little to no fine-tuning. However, for medical images, like chest X-rays, which are characterized by complex anatomical structures and diverse clinical conditions, there arises a need for representation learning techniques that can encode fine-grained details while preserving the broader contextual information. In this context, we introduce MLVICX (Multi-Level Variance-Covariance Exploration for Chest X-ray Self-Supervised Representation Learning), an approach to capture rich representations in the form of embeddings from chest X-ray images. Central to our approach is a novel multi-level variance and covariance exploration strategy that empowers the model to detect diagnostically meaningful patterns while reducing redundancy effectively. By enhancing the variance and covariance of the learned embeddings, MLVICX promotes the retention of critical medical insights by adapting both global and local contextual details. We demonstrate the performance of MLVICX in advancing self-supervised chest X-ray representation learning through comprehensive experiments. The performance enhancements we observe across various downstream tasks highlight the significance of the proposed approach in enhancing the utility of chest X-ray embeddings for precision medical diagnosis and comprehensive image analysis. For pertaining, we used the NIH-Chest X-ray dataset, while for downstream tasks, we utilized NIH-Chest X-ray, Vinbig-CXR, RSNA pneumonia, and SIIM-ACR Pneumothorax datasets. Overall, we observe more than 3% performance gains over SOTA SSL approaches in various downstream tasks.

Effective solutions for intelligent data collection in terrestrial cellular networks are crucial, especially in the context of Internet of Things applications. The limited spectrum and coverage area of terrestrial base stations pose challenges in meeting the escalating data rate demands of network users. Unmanned aerial vehicles, known for their high agility, mobility, and flexibility, present an alternative means to offload data traffic from terrestrial BSs, serving as additional access points. This paper introduces a novel approach to efficiently maximize the utilization of multiple UAVs for data traffic offloading from terrestrial BSs. Specifically, the focus is on maximizing user association with UAVs by jointly optimizing UAV trajectories and users association indicators under quality of service constraints. Since, the formulated UAVs control problem is nonconvex and combinatorial, this study leverages the multi agent reinforcement learning framework. In this framework, each UAV acts as an independent agent, aiming to maintain inter UAV cooperative behavior. The proposed approach utilizes the finite state Markov decision process to account for UAVs velocity constraints and the relationship between their trajectories and state space. A low complexity distributed state action reward state action algorithm is presented to determine UAVs optimal sequential decision making policies over training episodes. The extensive simulation results validate the proposed analysis and offer valuable insights into the optimal UAV trajectories. The derived trajectories demonstrate superior average UAV association performance compared to benchmark techniques such as Q learning and particle swarm optimization., Comment: Significant modification required to get novel results

In this paper, we study underlay device-to-device (D2D) communication systems empowered by a reconfigurable intelligent surface (RIS) for cognitive cellular networks. Considering Rayleigh fading channels and the general case where there exist both the direct and RIS-enabled D2D channels, the outage probability (OP) of the D2D communication link is presented in closed-form. Next, for the considered RIS-empowered underlaid D2D system, we frame an OP minimization problem. We target the joint optimization of the transmit power at the D2D source and the RIS placement, under constraints on the transmit power at the D2D source and on the limited interference imposed on the cellular user for two RIS deployment topologies. Due to the coupled optimization variables, the formulated optimization problem is extremely intractable. We propose an equivalent transformation which we are able to solve analytically. In the transformed problem, an expression for the average value of the signal-to-interference-noise ratio (SINR) at the D2D receiver is derived in closed-form. Our theoretical derivations are corroborated through simulation results, and various system design insights are deduced. It is indicatively showcased that the proposed RIS-empowered underlaid D2D system design outperforms the benchmark semi-adaptive optimal power and optimal distance schemes, offering $44\%$ and $20\%$ performance improvement, respectively.

This paper focuses on secure backscatter transmission in the presence of a passive multi-antenna eavesdropper through a symbiotic radio (SR) network. Specifically, a single-antenna backscatter device (BD) aims to transmit confidential information to a primary receiver (PR) by using a multi-antenna primary transmitter's (PT) signal, where the received symbols are jointly decoded at the PR. Our objective is to achieve confidential communications for BD while ensuring that the primary system's quality of service (QoS) requirements are met. We propose an alternating optimisation algorithm that maximises the achievable secrecy rate of BD by jointly optimising primary transmit beamforming and power sharing between information and artificial noise (AN) signals. Numerical results verify our analytical claims on the optimality of the proposed solution and the proposed methodology's underlying low complexity. Additionally, our simulations provide nontrivial design insights into the critical system parameters and quantify the achievable gains over the relevant benchmark schemes.

Orthogonal time frequency space (OTFS) has been widely acknowledged as a promising wireless technology for challenging transmission scenarios, including high-mobility channels. In this paper, we investigate the pilot design for the multi-user OTFS system based on the a priori statistical channel state information (CSI), where the practical threshold-based estimation scheme is adopted. Specifically, we first derive the a posteriori Cramer-Rao bound (PCRB) based on a priori channel information for each user. According to our derivation, the PCRB only relates to the user's pilot signal-to-noise ratio (SNR) and the range of delay and Doppler shifts under the practical power-delay and power-Doppler profiles. Then, a pilot scheme is proposed to minimize the average PCRB of different users, where a closed-form global optimal pilot power allocation is derived. Our numerical results verify the multi-user PCRB analysis. Also, we demonstrate an around 3 dB improvement in the average normalized-mean-square error (NMSE) by using the proposed pilot design in comparison to the conventional embedded pilot design under the same total pilot power.

Intelligent reflecting surface (IRS) has recently emerged as a promising technology for beyond fifth-generation (B5G) and 6G networks conceived from metamaterials that smartly tunes the signal reflections via a large number of low-cost passive reflecting elements. However, the IRS-assisted communication model and the optimization of available resources needs to be improved further for more efficient communications. This paper investigates the enhancement of received power at the user end in an IRS assisted wireless communication by jointly optimizing the phase shifts at the IRS elements and its location. Employing the conventional Friss transmission model, the relationship between the transmitted power and reflected power is established. The expression of received power incorporates the free space loss, reflection loss factor, physical dimension of the IRS panel, and radiation pattern of the transmit signal. Also, the expression of reflection coefficient of IRS panel is obtained by exploiting the existing data of radar communications. Initially exploring a single IRS element within a two-ray reflection model, we extend it to a more complex multi-ray reflection model with multiple IRS elements in 3D Cartesian space. The received power expression is derived in a more tractable form, then, it is maximized by jointly optimizing the underlying underlying variables, the IRS location and the phase shifts. To realize the joint optimization of underlying variables, first, the phase shifts of the IRS elements are optimized to achieve constructive interference of received signal components at the user. Subsequently, the location of the IRS is optimized at the obtained optimal phase shifts. Numerical insights and performance comparison reveal that joint optimization leads to a substantial 37% enhancement in received power compared to the closest competitive benchmark scheme.

Non-orthogonal multiple access (NOMA) has been recognized as one of the key enabling technologies for future generation wireless networks. Sharing the same time-frequency resource among users imposes secrecy challenges in NOMA in the presence of untrusted users. This paper characterizes the impact of user-pair selection on the secrecy performance of an untrusted NOMA system. In this regard, an optimization problem is formulated to maximize the secrecy rate of the strong user while satisfying the quality of service (QoS) demands of the user with poorer channel conditions. To solve this problem, we first obtain optimal power allocation in a two-user NOMA system, and then investigate the user-pair selection problem in a more generalized four user NOMA system. Extensive performance evaluations are conducted to validate the accuracy of the proposed results and present valuable insights on the impact of various system parameters on the secrecy performance of the NOMA communication system.

Non-orthogonal multiple access (NOMA) has come to the fore as a spectral-efficient technique for fifth-generation and beyond communication networks. We consider the downlink of a NOMA system with untrusted users. In order to consider a more realistic scenario, imperfect successive interference cancellation is assumed at the receivers during the decoding process. Since pair outage probability (POP) ensures a minimum rate guarantee to each user, it behaves as a measure of the quality of service for the pair of users. With the objective of designing a reliable communication protocol, we derive the closed-form expression of POP. Further, we find the optimal power allocation that minimizes the POP. Lastly, numerical results have been presented which validate the exactness of the analysis, and reveal the effect of various key parameters on achieved pair outage performance. In addition, we benchmark optimal power allocation against equal and fixed power allocations with respect to POP. The results indicate that optimal power allocation results in improved communication reliability., Comment: 6 pages, 5 figures, WCNC 2022

Non-orthogonal multiple access (NOMA) has come to the fore as a spectrally efficient technique for fifth-generation networks and beyond. At the same time, NOMA faces severe security issues in the presence of untrusted users due to successive interference cancellation (SIC)-based decoding at receivers. In this paper, to make the system model more realistic, we consider the impact of imperfect SIC during the decoding process. Assuming the downlink mode, we focus on designing a secure NOMA communication protocol for the considered system model with two untrusted users. In this regard, we obtain the power allocation bounds to achieve a positive secrecy rate for both near and far users. Analytical expressions of secrecy outage probability (SOP) for both users are derived to analyze secrecy performance. Closed-form approximations of SOPs are also provided to gain analytical insights. Lastly, numerical results have been presented, which validate the exactness of the analysis and reveal the effect of various key parameters on achieved secrecy performance., Comment: 5 pages, 5 figures, and EUSIPCO 2021

Non-orthogonal multiple access (NOMA) serves multiple users simultaneously via the same resource block by exploiting superposition coding at the transmitter and successive interference cancellation (SIC) at the receivers. Under practical considerations, perfect SIC may not be achieved. Thus, residual interference (RI) occurs inevitably due to imperfect SIC. In this work, we first propose a novel model for characterizing RI to provide a more realistic secrecy performance analysis of a downlink NOMA system under imperfect SIC at receivers. In the presence of untrusted users, NOMA has an inherent security flaw. Therefore, for this untrusted users' scenario, we derive new analytical expressions of secrecy outage probability (SOP) for each user in a two-user untrusted NOMA system by using the proposed RI model. To further shed light on the obtained results and obtain a deeper understanding, a high signal-to-noise ratio approximation of the SOPs is also obtained. Lastly, numerical investigations are provided to validate the accuracy of the desired analytical results and present valuable insights into the impact of various system parameters on the secrecy rate performance of the secure NOMA communication system., Comment: 6 pages, 5 figures, GLOBECOM 2021

The security of confidential information associated with devices in the industrial Internet of Things (IIoT) network is a serious concern. This article focuses on achieving a nonorthogonal multiple access (NOMA)-enabled secure IIoT network in the presence of untrusted devices by jointly optimizing the resources, such as decoding order and power allocated to devices. Assuming that the devices are resource-constrained for performing perfect successive interference cancellation (SIC), we characterize the residual interference at receivers with the linear model. Firstly, considering all possible decoding orders in an untrusted scenario, we obtain secure decoding orders that are feasible to obtain a positive secrecy rate for each device. Then, under the secrecy fairness criterion, we formulate a joint optimization problem of maximizing the minimum secrecy rate among devices. Since the formulated problem is non-convex and combinatorial, we first obtain the optimal secure decoding order and then solve it for power allocation by analyzing Karush-Kuhn-Tucker points. Thus, we provide the closed-form global-optimal solution of the formulated optimization problem. Numerical results validate the analytical claims and demonstrate an interesting observation that the conventional decoding order and assigning more power allocation to the weak device, as presumed in many works on NOMA, is not an optimal strategy from the secrecy fairness viewpoint. Also, the average percentage gain of about 22.75%, 50.58%, 94.59%, and 98.16%, respectively, is achieved by jointly optimized solution over benchmarks ODEP (optimal decoding order, equal power allocation), ODFP (optimal decoding order, fixed power allocation), FDEP (fixed decoding order, equal power allocation), and FDFP (fixed decoding order, fixed power allocation)., Comment: 10 pages and 6 figures

Large pre-trained models, also known as foundation models (FMs), are trained in a task-agnostic manner on large-scale data and can be adapted to a wide range of downstream tasks by fine-tuning, few-shot, or even zero-shot learning. Despite their successes in language and vision tasks, we have yet seen an attempt to develop foundation models for geospatial artificial intelligence (GeoAI). In this work, we explore the promises and challenges of developing multimodal foundation models for GeoAI. We first investigate the potential of many existing FMs by testing their performances on seven tasks across multiple geospatial subdomains including Geospatial Semantics, Health Geography, Urban Geography, and Remote Sensing. Our results indicate that on several geospatial tasks that only involve text modality such as toponym recognition, location description recognition, and US state-level/county-level dementia time series forecasting, these task-agnostic LLMs can outperform task-specific fully-supervised models in a zero-shot or few-shot learning setting. However, on other geospatial tasks, especially tasks that involve multiple data modalities (e.g., POI-based urban function classification, street view image-based urban noise intensity classification, and remote sensing image scene classification), existing foundation models still underperform task-specific models. Based on these observations, we propose that one of the major challenges of developing a FM for GeoAI is to address the multimodality nature of geospatial tasks. After discussing the distinct challenges of each geospatial data modality, we suggest the possibility of a multimodal foundation model which can reason over various types of geospatial data through geospatial alignments. We conclude this paper by discussing the unique risks and challenges to develop such a model for GeoAI.

Knowledge distillation is a common technique for improving the performance of a shallow student network by transferring information from a teacher network, which in general, is comparatively large and deep. These teacher networks are pre-trained and often uncalibrated, as no calibration technique is applied to the teacher model while training. Calibration of a network measures the probability of correctness for any of its predictions, which is critical in high-risk domains. In this paper, we study how to obtain a calibrated student from an uncalibrated teacher. Our approach relies on the fusion of the data-augmentation techniques, including but not limited to cutout, mixup, and CutMix, with knowledge distillation. We extend our approach beyond traditional knowledge distillation and find it suitable for Relational Knowledge Distillation and Contrastive Representation Distillation as well. The novelty of the work is that it provides a framework to distill a calibrated student from an uncalibrated teacher model without compromising the accuracy of the distilled student. We perform extensive experiments to validate our approach on various datasets, including CIFAR-10, CIFAR-100, CINIC-10 and TinyImageNet, and obtained calibrated student models. We also observe robust performance of our approach while evaluating it on corrupted CIFAR-100C data.

A large number of unidentified sources found by astronomical surveys and other observations necessitate the use of an automated classification technique based on machine learning methods. The aim of this paper is to find a suitable automated classifier to identify the point X-ray sources in the Chandra Source Catalogue (CSC) 2.0 in the categories of active galactic nuclei (AGN), X-ray emitting stars, young stellar objects (YSOs), high-mass X-ray binaries (HMXBs), low-mass X-ray binaries (LMXBs), ultra luminous X-ray sources (ULXs), cataclysmic variables (CVs), and pulsars. The catalogue consists of approx 3,17,000 sources, out of which we select 2,77,069 point sources based on the quality flags available in CSC 2.0. In order to identify unknown sources of CSC 2.0, we use multi-wavelength features, such as magnitudes in optical/UV bands from Gaia-EDR3, SDSS and GALEX, and magnitudes in IR bands from 2MASS, WISE and MIPS-Spitzer, in addition to X-ray features (flux and variability) from CSC 2.0. We find the Light Gradient Boosted Machine, an advanced decision tree-based machine learning classification algorithm, suitable for our purpose and achieve $93\%$ precision, $93\%$ recall score and 0.91 Mathew's Correlation coefficient score. With the trained classifier, we identified 54,770 (14,066) sources with more than $3{\sigma}$ (4${\sigma}$) confidence, out of which there are 32,600 (8,574) AGNs, 16,148 (5,166) stars, 5,184 (208) YSOs, 439 (46) HMXBs, 197 (71) LMXBs, 50 (0) ULXs, 89 (1) CVs, and 63 (0) pulsars. This method can also be useful for identifying sources of other catalogues reliably., Comment: 12 pages, 6 figures, 9 tables, accepted for publication in Monthly Notices of the Royal Astronomical Society