Recent Submissions

• EMF-Aware Cellular Networks in RIS-Assisted Environments

(2021-01-21) [Preprint]
The dense deployment of the 5th-generation cellular networks (5G) and beyond has triggered health concerns due to the electric and magnetic fields (EMF) exposure. In this paper, we propose minimizing the populations’ exposure to EMF by considering a smart radio environment with a reconfigurable intelligent surface (RIS). We propose a novel algorithm for the RIS phase design that minimizes an exposure index (EI), in terms of the specific absorption rate (SAR), while maintaining a minimum target quality of service (QoS). The proposed algorithm achieves up to 60% and 50% reduction in EI compared to schemes without RISs and with non-optimized RIS, respectively.
• Protein Phosphatase 1 regulates atypical chromosome segregation and cell polarity during mitotic and meiotic division in Plasmodium sexual stages

(Cold Spring Harbor Laboratory, 2021-01-17) [Preprint]
AbstractPP1 is a conserved eukaryotic serine/threonine phosphatase that regulates many aspects of mitosis and meiosis, often working in concert with other phosphatases, such as CDC14 and CDC25. The proliferative stages of the parasite life cycle include sexual development within the mosquito vector, with male gamete formation characterized by an atypical rapid mitosis, consisting of three rounds of DNA synthesis, successive spindle formation with clustered kinetochores, and a meiotic stage during zygote to ookinete development following fertilization. It is unclear how PP1 is involved in these unusual processes. Using real-time live-cell and ultrastructural imaging, conditional gene knockdown, RNA-seq and proteomic approaches, we show that Plasmodium PP1 is involved in both chromosome segregation during mitotic exit, and establishment of cell polarity during zygote development in the mosquito midgut, suggesting that small molecule inhibitors of PP1 should be explored for blocking parasite transmission.
• Change-point detection using spectral PCA for multivariate time series

(arXiv, 2021-01-12) [Preprint]
We propose a two-stage approach Spec PC-CP to identify change points in multivariate time series. In the first stage, we obtain a low-dimensional summary of the high-dimensional time series by Spectral Principal Component Analysis (Spec-PCA). In the second stage, we apply cumulative sum-type test on the Spectral PCA component using a binary segmentation algorithm. Compared with existing approaches, the proposed method is able to capture the lead-lag relationship in time series. Our simulations demonstrate that the Spec PC-CP method performs significantly better than competing methods for detecting change points in high-dimensional time series. The results on epileptic seizure EEG data and stock data also indicate that our new method can efficiently {detect} change points corresponding to the onset of the underlying events.
• MAAS: Multi-modal Assignation for Active Speaker Detection

(arXiv, 2021-01-11) [Preprint]
Active speaker detection requires a solid integration of multi-modal cues. While individual modalities can approximate a solution, accurate predictions can only be achieved by explicitly fusing the audio and visual features and modeling their temporal progression. Despite its inherent muti-modal nature, current methods still focus on modeling and fusing short-term audiovisual features for individual speakers, often at frame level. In this paper we present a novel approach to active speaker detection that directly addresses the multi-modal nature of the problem, and provides a straightforward strategy where independent visual features from potential speakers in the scene are assigned to a previously detected speech event. Our experiments show that, an small graph data structure built from a single frame, allows to approximate an instantaneous audio-visual assignment problem. Moreover, the temporal extension of this initial graph achieves a new state-of-the-art on the AVA-ActiveSpeaker dataset with a mAP of 88.8\%.
• Dark Self-Healing Mediated Negative Photoconductivity of Lead-Free Cs3Bi2Cl9 Perovskite Single Crystal

(arXiv, 2021-01-07) [Preprint]
Halide perovskites are recently emerged as one of the frontline optoelectronic materials for device applications and have been extensively studied in past few years. Among these while, lead-based materials were most widely explored, investigation of optical properties of lead-free perovskites is limited. Being optically active, these materials were expected to show light-induced enhanced photoconductivity and the same was reported for lead halide perovskite single crystals. However, on contrary, herein, light-induced degradation of bismuth halide perovskite Cs3Bi2Cl9 single crystals is reported which was evidenced by negative photoconductivity with slow recovery. The femtosecond transient reflectance (fs-TR) spectroscopy studies further revealed these electronic transport properties were due to the formation of light-activated metastable trap states within the perovskite crystal. The figure of merits of Cs3Bi2Cl9 single-crystal detectors such as responsivity (17 mA/W), detectivity (6.23 X 10power 11 Jones) and the ratio of current in dark to light (~7160) was calculated and it is found that they are comparable or higher to reported perovskite single crystals based positive photodetectors. This observation for lead-free perovskite single crystals which were optically active but showed retroactive photocurrent on irradiation remained unique for such materials.
• Landslide size matters: a new spatial predictive paradigm

(California Digital Library (CDL), 2021-01-07) [Preprint]
The standard definition of landslide hazard requires the estimation of where, when (or how frequently) and how large a given landslide event may be. The geomorphological community involved in statistical models has addressed the component pertaining to how large a landslide event may be by introducing the concept of landslide-event magnitude scale. This scale, which depends on the planimetric area of the given population of landslides, in analogy to the earthquake magnitude, has been expressed with a single value per landslide event. As a result, the geographic or spatially-distributed estimation of how large a population of landslide may be when considered at the slope scale, has been disregarded in statistically-based landslide hazard studies. Conversely, the estimation of the landslide extent has been commonly part of physically-based applications, though their implementation is often limited to very small regions.
• Tractable Bayes of Skew-Elliptical Link Models for Correlated Binary Data

(arXiv, 2021-01-06) [Preprint]
Correlated binary response data with covariates are ubiquitous in longitudinal or spatial studies. Among the existing statistical models the most well-known one for this type of data is the multivariate probit model, which uses a Gaussian link to model dependence at the latent level. However, a symmetric link may not be appropriate if the data are highly imbalanced. Here, we propose a multivariate skew-elliptical link model for correlated binary responses, which includes the multivariate probit model as a special case. Furthermore, we perform Bayesian inference for this new model and prove that the regression coefficients have a closed-form unified skew-elliptical posterior. The new methodology is illustrated by application to COVID-19 pandemic data from three different counties of the state of California, USA. By jointly modeling extreme spikes in weekly new cases, our results show that the spatial dependence cannot be neglected. Furthermore, the results also show that the skewed latent structure of our proposed model improves the flexibility of the multivariate probit model and provides better fit to our highly imbalanced dataset.
• Ion-exchange doped polymers at the degenerate limit: what limits conductivity at 100% doping efficiency?

(arXiv, 2021-01-05) [Preprint]
Doping of semiconducting polymers has seen a surge in research interest driven by emerging applications in sensing, bioelectronics and thermoelectrics. A recent breakthrough was a doping technique based on ion-exchange, which separates the redox and charge compensation steps of the doping process. The improved microstructural control this process allows enables us for the first time to systematically address a longstanding but still poorly understood question: what limits the electrical conductivity at high doping levels? Is it the formation of charge carrier traps in the Coulomb potentials of the counterions, or is it the structural disorder in the polymer lattice? Here, we apply ion-exchange doping to several classes of high mobility conjugated polymers and identify experimental conditions that achieve near 100% doping efficiency under degenerate conditions with nearly 1 charge per monomer. We demonstrate very high conductivities up to 1200 S/cm in semicrystalline polymer systems, and show that in this regime conductivity is poorly correlated with ionic size, but strongly correlated with paracrystalline disorder. This observation, backed by a detailed electronic structure model that incorporates ion-hole and hole-hole interactions and a carefully parameterized model of disorder, indicates that trapping by dopant ions is negligible, and that maximizing crystalline order is critical to improving conductivity.
• A simple approach to proving the existence, uniqueness, and strong and weak convergence rates for a broad class of McKean--Vlasov equations

(arXiv, 2021-01-04) [Preprint]
By employing a system of interacting stochastic particles as an approximation of the McKean--Vlasov equation and utilizing classical stochastic analysis tools, namely It\^o's formula and Kolmogorov--Chentsov continuity theorem, we prove the existence and uniqueness of strong solutions for a broad class of McKean--Vlasov equations. Considering an increasing number of particles in the approximating stochastic particle system, we also prove the $L^p$ strong convergence rate and derive the weak convergence rates using the Kolmogorov backward equation and variations of the stochastic particle system. Our convergence rates were verified by numerical experiments which also indicate that the assumptions made here and in the literature can be relaxed.
• A Survey on Integrated Access and Backhaul Networks

(arXiv, 2021-01-04) [Preprint]
Benefiting from the usage of the high-frequency band, utilizing part of the large available bandwidth for wireless backhauling is feasible without considerable performance sacrifice. In this context, integrated access and backhaul (IAB) was proposed by 3GPP to reduce the fiber optics deployment cost of 5G and beyond networks. In this paper, we first give a brief introduction of IAB based on the 3GPP release. After that, we survey existing research on IAB networks, the integrations of IAB to cache-enabled network, optical communication transport network, and the non-terrestrial network. Finally, we discuss the challenges and opportunities that might arise while developing and commercializing IAB networks.
• Using BART for Multiobjective Optimization of Noisy Multiple Objectives

(arXiv, 2021-01-04) [Preprint]
Techniques to reduce the energy burden of an Industry 4.0 ecosystem often require solving a multiobjective optimization problem. However, collecting experimental data can often be either expensive or time-consuming. In such cases, statistical methods can be helpful. This article proposes Pareto Front (PF) and Pareto Set (PS) estimation methods using Bayesian Additive Regression Trees (BART), which is a non-parametric model whose assumptions are typically less restrictive than popular alternatives, such as Gaussian Processes. The performance of our BART-based method is compared to a GP-based method using analytic test functions, demonstrating convincing advantages. Finally, our BART-based methodology is applied to a motivating Industry 4.0 engineering problem.
• CIZSL++: Creativity Inspired Generative Zero-Shot Learning

(arXiv, 2021-01-01) [Preprint]
Zero-shot learning (ZSL) aims at understanding unseen categories with no training examples from class-level descriptions. To improve the discriminative power of ZSL, we model the visual learning process of unseen categories with inspiration from the psychology of human creativity for producing novel art. First, we propose CIZSL-v1 as a creativity inspired model for generative ZSL. We relate ZSL to human creativity by observing that ZSL is about recognizing the unseen, and creativity is about creating a likable unseen. We introduce a learning signal inspired by creativity literature that explores the unseen space with hallucinated class-descriptions and encourages careful deviation of their visual feature generations from seen classes while allowing knowledge transfer from seen to unseen classes. Second, CIZSL-v2 is proposed as an improved version of CIZSL-v1 for generative zero-shot learning. CIZSL-v2 consists of an investigation of additional inductive losses for unseen classes along with a semantic guided discriminator. Empirically, we show consistently that CIZSL losses can improve generative ZSL models on the challenging task of generalized ZSL from a noisy text on CUB and NABirds datasets. We also show the advantage of our approach to Attribute-based ZSL on AwA2, aPY, and SUN datasets. We also show that CIZSL-v2 has improved performance compared to CIZSL-v1.
• Virtual element approximation of eigenvalue problems

(arXiv, 2020-12-31) [Preprint]
We discuss the approximation of eigenvalue problems associated with elliptic partial differential equations using the virtual element method. After recalling the abstract theory, we present a model problem, describing in detail the features of the scheme, and highligting the effects of the stabilizing parameters. We conlcude the discussion with a survey of several application examples.
• SALA: Soft Assignment Local Aggregation for 3D Semantic Segmentation

(arXiv, 2020-12-29) [Preprint]
We introduce the idea of using learnable neighbor-togrid soft assignment in grid-based aggregation functions for the task of 3D semantic segmentation. Previous methods in literature operate on a predefined geometric grid such as local volume partitions or irregular kernel points. These methods use geometric functions to assign local neighbors to their corresponding grid. Such geometric heuristics are potentially sub-optimal for the end task of semantic segmentation. Furthermore, they are applied uniformly throughout the depth of the network. A more general alternative would allow the network to learn its own neighbor-to-grid assignment function that best suits the end task. Since it is learnable, this mapping has the flexibility to be different per layer. This paper leverages learned neighbor-to-grid soft assignment to define an aggregation function that balances efficiency and performance. We demonstrate the efficacy of our method by reaching state-of-the-art (SOTA) performance on S3DIS with almost 10× less parameters than the current reigning method. We also demonstrate competitive performance on ScanNet and PartNet as compared with much larger SOTA models.
• Single-cell Individual Complete mtDNA Sequencing Uncovers Hidden Mitochondrial Heterogeneity in Human and Mouse Oocytes

(Cold Spring Harbor Laboratory, 2020-12-29) [Preprint]
The ontogeny and dynamics of mtDNA heteroplasmy remain unclear due to limitations of current mtDNA sequencing methods. We developed individual Mitochondrial Genome sequencing (iMiGseq) of full-length mtDNA for ultra-sensitive variant detection, complete haplotyping, and unbiased evaluation of heteroplasmy levels, all at the individual mtDNA molecule level. iMiGseq uncovers unappreciated levels of heteroplasmic variants in single healthy human oocytes well below the current 1% detection limit, of which numerous variants are detrimental and could contribute to late-onset mitochondrial disease and cancer. Extreme mtDNA heterogeneity among oocytes of the same mouse female, and a strong selection against deleterious mutations in human oocytes are observed. iMiGseq could comprehensively characterize and haplotype single-nucleotide and structural variants of mtDNA and their genetic linkage in NARP/Leigh syndrome patient-derived cells. Therefore, iMiGseq could not only elucidate the mitochondrial etiology of diseases, but also help diagnose and prevent mitochondrial diseases with unprecedented precision.
• The Relaxation Limit of Bipolar Fluid Models

(arXiv, 2020-12-28) [Preprint]
This work establishes the relaxation limit from a bipolar Euler-Poisson system with friction towards a bipolar drift-diffusion system. A weak-strong formalism is developed and, within this framework, a dissipative weak solution of the bipolar Euler-Poisson system converges in the high-friction regime to a conservative, bounded away from vacuum, strong solution of the bipolar drift-diffusion system. This limiting process is based on a relative entropy identity for the bipolar fluid system.
• Energy-conserving 3D elastic wave simulation with finite difference discretization on staggered grids with nonconforming interfaces

(arXiv, 2020-12-27) [Preprint]
In this work, we describe an approach to stably simulate the 3D isotropic elastic wave propagation using finite difference discretization on staggered grids with nonconforming interfaces. Specifically, we consider simulation domains composed of layers of uniform grids with different grid spacings, separated by planar interfaces. This discretization setting is motivated by the observation that wave speeds of earth media tend to increase with depth due to sedimentation and consolidation processes. We demonstrate that the layer-wise finite difference discretization approach has the potential to significantly reduce the simulation cost, compared to its counterpart that uses holistically uniform grids. Such discretizations are enabled by summation-by-parts finite difference operators, which are standard finite difference operators with special adaptations near boundaries or interfaces, and simultaneous approximation terms, which are penalty terms appended to the discretized system to weakly impose boundary or interface conditions. Combined with specially designed interpolation operators, the discretized system is shown to preserve the energy-conserving property of the continuous elastic wave equation, and a fortiori ensure the stability of the simulation. Numerical examples are presented to corroborate these analytical developments.
• The gap-free rice genomes provide insights for centromere structure and function exploration and graph-based pan-genome construction

(Cold Spring Harbor Laboratory, 2020-12-25) [Preprint]
Asia rice (Oryza sativa) is divided into two subgroups, indica/xian and japonica/geng, the former has greater intraspecific diversity than the latter. Here, for the first time, we report the assemblies and analyses of two gap-free xian rice varieties Zhenshan 97 (ZS97) and Minghui 63 (MH63). Genomic sequences of these elite hybrid parents express extensive difference as the foundation for studying heterosis. Furthermore, the gap-free rice genomes provide global insights to investigate the structure and function of centromeres in different chromosomes. All the rice centromeric regions share conserved centromere-specific satellite motifs but with different copy numbers and structures. Importantly, we show that there are >1,500 genes in centromere regions and ~16% of them are actively expressed. Based on MH63 gap-free reference genome, a graph-based rice pan-genome (Os-GPG) was constructed containing presence/absence variations of 79 rice varieties. Compared with the other rice varieties, MH63 contained the largest number of resistance genes. The acquisition of ZS97 and MH63 gap-free genomes and graph-based pan-genome of rice lays a solid foundation for the study of genome structure and function in plants.
• Three-dimensional Electromagnetic Void Space

(arXiv, 2020-12-25) [Preprint]
We report a realization of three-dimensional (3D) electromagnetic void space. Despite occupying a finite volume of space, such a medium is optically equivalent to an infinitesimal point where electromagnetic waves experience no phase accumulation. The 3D void space is realized by constructing all-dielectric 3D photonic crystals such that the effective permittivity and permeability vanish simultaneously, forming a six-fold Dirac-like point with Dirac-like linear dispersions at the center of the Brillouin Zone. We demonstrate, both theoretically and experimentally, that such a 3D void space exhibits unique properties and rich functionalities absent in any other electromagnetic media, such as boundary-control transmission switching and 3D perfect wave-steering mechanisms. Especially, contrary to the photonic "doping" effect in its two-dimensional counterpart, the 3D void space exhibits an amazing property of "impurity-immunity". Our work paves a road towards the realization of 3D void space where electromagnetic waves can be manipulated in unprecedented ways.
• Interpolating Points on a Non-Uniform Grid using a Mixture of Gaussians

(arXiv, 2020-12-24) [Preprint]
In this work, we propose an approach to perform non-uniform image interpolation based on a Gaussian Mixture Model. Traditional image interpolation methods, like nearest neighbor, bilinear, Hamming, Lanczos, etc. assume that the coordinates you want to interpolate from, are positioned on a uniform grid. However, it is not always the case in practice and we develop an interpolation method that is able to generate an image from arbitrarily positioned pixel values. We do this by representing each known pixel as a 2D normal distribution and considering each output image pixel as a sample from the mixture of all the known ones. Apart from the ability to reconstruct an image from arbitrarily positioned set of pixels, this also allows us to differentiate through the interpolation procedure, which might be helpful for downstream applications. Our optimized CUDA kernel and the source code to reproduce the benchmarks is located at https://github.com/universome/non-uniform-interpolation.