arXiv subjects

Xiao Hu

Publications and source records attributed to Xiao Hu.

At least 181 records · Page 10Linked to original sources

arXiv2022

Why Accuracy Is Not Enough: The Need for Consistency in Object Detection

Object detectors are vital to many modern computer vision applications. However, even state-of-the-art object detectors are not perfect. On two images that look similar to human eyes, the same detector can make different predictions because of small image distortions like camera sensor noise and lighting changes. This problem is called inconsistency. Existing accuracy metrics do not properly account for inconsistency, and similar work in this area only targets improvements on artificial image distortions. Therefore, we propose a method to use non-artificial video frames to measure object detection consistency over time, across frames. Using this method, we show that the consistency of modern object detectors ranges from 83.2% to 97.1% on different video datasets from the Multiple Object Tracking Challenge. We conclude by showing that applying image distortion corrections like .WEBP Image Compression and Unsharp Masking can improve consistency by as much as 5.1%, with no loss in accuracy.

cs.CV↗

arXiv2022

Irrelevant Pixels are Everywhere: Find and Exclude Them for More Efficient Computer Vision

Computer vision is often performed using Convolutional Neural Networks (CNNs). CNNs are compute-intensive and challenging to deploy on power-contrained systems such as mobile and Internet-of-Things (IoT) devices. CNNs are compute-intensive because they indiscriminately compute many features on all pixels of the input image. We observe that, given a computer vision task, images often contain pixels that are irrelevant to the task. For example, if the task is looking for cars, pixels in the sky are not very useful. Therefore, we propose that a CNN be modified to only operate on relevant pixels to save computation and energy. We propose a method to study three popular computer vision datasets, finding that 48% of pixels are irrelevant. We also propose the focused convolution to modify a CNN's convolutional layers to reject the pixels that are marked irrelevant. On an embedded device, we observe no loss in accuracy, while inference latency, energy consumption, and multiply-add count are all reduced by about 45%.

cs.CV↗

arXiv2022

Formation of Dust Rings and Gaps in Non-ideal MHD Disks Through Meridional Gas Flows

Rings and gaps are commonly observed in the dust continuum emission of young stellar disks. Previous studies have shown that substructures naturally develop in the weakly ionized gas of magnetized, non-ideal MHD disks. The gas rings are expected to trap large mm/cm-sized grains through pressure gradient-induced radial dust-gas drift. Using 2D (axisymmetric) MHD simulations that include ambipolar diffusion and dust grains of three representative sizes (1~mm, 3.3~mm, and 1~cm), we show that the grains indeed tend to drift radially relative to the gas towards the centers of the gas rings, at speeds much higher than in a smooth disk because of steeper pressure gradients. However, their spatial distribution is primarily controlled by meridional gas motions, which are typically much faster than the dust-gas drift. In particular, the grains that have settled near the midplane are carried rapidly inwards by a fast accretion stream to the inner edges of the gas rings, where they are lifted up by the gas flows diverted away from the midplane by a strong poloidal magnetic field. The flow pattern in our simulation provides an attractive explanation for the meridional flows recently inferred in HD 163296 and other disks, including both "collapsing" regions where the gas near the disk surface converges towards the midplane and a disk wind. Our study highlights the prevalence of the potentially observable meridional flows associated with the gas substructure formation in non-ideal MHD disks and their crucial role in generating rings and gaps in dust.

astro-ph.EP↗

arXiv2022

Higgs-Leggett mechanism for the elusive 6e superconductivity observed in Kagome vanadium-based superconductors

A recent Little-Parks experiment on Kagome-structured superconductor CsV_3Sb_5 demonstrated remarkable resistance oscillations with period \phi0/3=hc/6e. Here, we perform analysis based on a theory involving three 2e superconductivity (SC) order parameters associated with the three reciprocal lattice vectors which connect M points of the hexagonal Brillouin zone. In a ring geometry we unveil that, as a series of metastable states, phase of one SC order parameter winds 2πmore or less than the other two ones around the ring, which yields local free energy minima at integer multiples of \phi0/3. Intriguingly, the ground-state degeneracy associated with a Z_2 chirality is crucial, and the Higgs-Leggett mechanism stabilizes domain walls (DW) between chiral domains. At low temperatures DW are expelled from the system resulting in free energy minima only at integer multiples of ϕ_0. Our theory explains successfully the 6e SC observed in experiments, which opens a door for approaching rich physics of Kagome superconductors.

cond-mat.supr-con↗

arXiv2022

Log-Spectral Matching GAN: PPG-based Atrial Fibrillation Detection can be Enhanced by GAN-based Data Augmentation with Integration of Spectral Loss

Photoplethysmography (PPG) is a ubiquitous physiological measurement that detects beat-to-beat pulsatile blood volume changes and hence has a potential for monitoring cardiovascular conditions, particularly in ambulatory settings. A PPG dataset that is created for a particular use case is often imbalanced, due to a low prevalence of the pathological condition it targets to predict and the paroxysmal nature of the condition as well. To tackle this problem, we propose log-spectral matching GAN (LSM-GAN), a generative model that can be used as a data augmentation technique to alleviate the class imbalance in a PPG dataset to train a classifier. LSM-GAN utilizes a novel generator that generates a synthetic signal without a up-sampling process of input white noises, as well as adds the mismatch between real and synthetic signals in frequency domain to the conventional adversarial loss. In this study, experiments are designed focusing on examining how the influence of LSM-GAN as a data augmentation technique on one specific classification task - atrial fibrillation (AF) detection using PPG. We show that by taking spectral information into consideration, LSM-GAN as a data augmentation solution can generate more realistic PPG signals. The code of LSM-GAN is available at https://github.com/chengding0713/Log-Spectral-matching-GAN.

eess.SP↗

arXiv2022

Ranked Enumeration of Join Queries with Projections

Join query evaluation with ordering is a fundamental data processing task in relational database management systems. SQL and custom graph query languages such as Cypher offer this functionality by allowing users to specify the order via the ORDER BY clause. In many scenarios, the users also want to see the first $k$ results quickly (expressed by the LIMIT clause), but the value of $k$ is not predetermined as user queries are arriving in an online fashion. Recent work has made considerable progress in identifying optimal algorithms for ranked enumeration of join queries that do not contain any projections. In this paper, we initiate the study of the problem of enumerating results in ranked order for queries with projections. Our main result shows that for any acyclic query, it is possible to obtain a near-linear (in the size of the database) delay algorithm after only a linear time preprocessing step for two important ranking functions: sum and lexicographic ordering. For a practical subset of acyclic queries known as star queries, we show an even stronger result that allows a user to obtain a smooth tradeoff between faster answering time guarantees using more preprocessing time. Our results are also extensible to queries containing cycles and unions. We also perform a comprehensive experimental evaluation to demonstrate that our algorithms, which are simple to implement, improve up to three orders of magnitude in the running time over state-of-the-art algorithms implemented within open-source RDBMS and specialized graph databases.

cs.DB↗

arXiv2021

Topological magnon modes on honeycomb lattice with coupling textures

Topological magnon modes are expected to be useful for novel applications such as robust information propagation, since they are immune to backscattering and robust against disorder. Although there are several of theoretical proposals for topological magnon modes and growing experimental efforts for realizing them by now, it is still desirable to add complementary insights on this important phenomenon. Here, we propose a new scheme to achieve topological magnon where only nearest-neighbour exchange couplings on honeycomb lattice are necessary. In both ferromagnets and antiferromagnets, tuning exchange couplings between and inside hexagonal unit cells induces a topological state accompanied by a band inversion between p-orbital and d-orbital like magnon modes. Topological magnon modes appear at the interface between a topological domain and a trivial domain with magnon currents, which counterpropagate depending on pseudospins originated from orbital angular momenta of magnon modes. This mimics the spin-momentum locking phenomenon in the quantum spin Hall effect.

cond-mat.mes-hall↗

arXiv2021

Efficient Computer Vision on Edge Devices with Pipeline-Parallel Hierarchical Neural Networks

Computer vision on low-power edge devices enables applications including search-and-rescue and security. State-of-the-art computer vision algorithms, such as Deep Neural Networks (DNNs), are too large for inference on low-power edge devices. To improve efficiency, some existing approaches parallelize DNN inference across multiple edge devices. However, these techniques introduce significant communication and synchronization overheads or are unable to balance workloads across devices. This paper demonstrates that the hierarchical DNN architecture is well suited for parallel processing on multiple edge devices. We design a novel method that creates a parallel inference pipeline for computer vision problems that use hierarchical DNNs. The method balances loads across the collaborating devices and reduces communication costs to facilitate the processing of multiple video frames simultaneously with higher throughput. Our experiments consider a representative computer vision problem where image recognition is performed on each video frame, running on multiple Raspberry Pi 4Bs. With four collaborating low-power edge devices, our approach achieves 3.21X higher throughput, 68% less energy consumption per device per frame, and 58% decrease in memory when compared with existing single-device hierarchical DNNs.

cs.CV↗

arXiv2021

Self-consistent ring model in protoplanetary disks: temperature dips and substructure formation

Rings and gaps are ubiquitous in protoplanetary disks. Larger dust grains will concentrate in gaseous rings more compactly due to stronger aerodynamic drag. However, the effects of dust concentration on the ring's thermal structure have not been explored. Using MCRT simulations, we self-consistently construct ring models by iterating the ring's thermal structure, hydrostatic equilibrium, and dust concentration. We set up rings with two dust populations having different settling and radial concentration due to their different sizes. We find two mechanisms that can lead to temperature dips around the ring. When the disk is optically thick, the temperature drops outside the ring, which is the shadowing effect found in previous works adopting a single-dust population in the disk. When the disk is optically thin, a second mechanism due to excess cooling of big grains is found. Big grains cool more efficiently, which leads to a moderate temperature dip within the ring where big dust resides. This dip is close to the center of the ring. Such temperature dip within the ring can lead to particle pile-up outside the ring and feedback to the dust distribution and thermal structure. We couple the MCRT calculations with a 1D dust evolution model and show that the ring evolves to a different shape and may even separate to several rings. Overall, dust concentration within rings has moderate effects on the disk's thermal structure, and self-consistent model is crucial not only for protoplanetary disk observations but also for planetesimal and planet formation studies.

astro-ph.EP↗

arXiv2021

Low-Power Multi-Camera Object Re-Identification using Hierarchical Neural Networks

Low-power computer vision on embedded devices has many applications. This paper describes a low-power technique for the object re-identification (reID) problem: matching a query image against a gallery of previously seen images. State-of-the-art techniques rely on large, computationally-intensive Deep Neural Networks (DNNs). We propose a novel hierarchical DNN architecture that uses attribute labels in the training dataset to perform efficient object reID. At each node in the hierarchy, a small DNN identifies a different attribute of the query image. The small DNN at each leaf node is specialized to re-identify a subset of the gallery: only the images with the attributes identified along the path from the root to a leaf. Thus, a query image is re-identified accurately after processing with a few small DNNs. We compare our method with state-of-the-art object reID techniques. With a 4% loss in accuracy, our approach realizes significant resource savings: 74% less memory, 72% fewer operations, and 67% lower query latency, yielding 65% less energy consumption.

cs.CV↗

arXiv2021

Topological Ring-Cavity Laser Formed by Honeycomb Photonic Crystals

We clarify theoretically that the topological ring-cavity (TRC) modes propagating along the interface between two honeycomb-type photonic crystals distinct in topology can be exploited for achieving stable single-mode lasing, with the maximal intensity larger than a whispering-gallery-mode counterpart by order of magnitude. Especially, we show that the TRC modes located at the bulk bandgap center benefit maximally from the gain profile since they are most concentrated and uniform along the ring cavity, and that, inheriting from the Dirac-like dispersion of topological interface states, they are separated in frequency from each other and from other photonic modes, both favoring intrinsically single-mode lasing. A TRC mode running in a specific direction with desired orbital angular momentum can be stimulated selectively by injecting circularly polarized light. The TRC laser proposed in the present work can be fabricated by means of advanced semiconductor nanotechnologies, which generates chiral laser beams ideal for novel photonic functions.

physics.optics↗

arXiv2021

Dynamic Enumeration of Similarity Joins

This paper considers enumerating answers to similarity-join queries under dynamic updates: Given two sets of $n$ points $A,B$ in $\mathbb{R}^d$, a metric $ϕ(\cdot)$, and a distance threshold $r > 0$, report all pairs of points $(a, b) \in A \times B$ with $ϕ(a,b) \le r$. Our goal is to store $A,B$ into a dynamic data structure that, whenever asked, can enumerate all result pairs with worst-case delay guarantee, i.e., the time between enumerating two consecutive pairs is bounded. Furthermore, the data structure can be efficiently updated when a point is inserted into or deleted from $A$ or $B$. We propose several efficient data structures for answering similarity-join queries in low dimension. For exact enumeration of similarity join, we present near-linear-size data structures for $\ell_1, \ell_\infty$ metrics with $\log^{O(1)} n$ update time and delay. We show that such a data structure is not feasible for the $\ell_2$ metric for $d \ge 4$. For approximate enumeration of similarity join, where the distance threshold is a soft constraint, we obtain a unified linear-size data structure for $\ell_p$ metric, with $\log^{O(1)} n$ delay and update time. In high dimensions, we present an efficient data structure with worst-case delay-guarantee using locality sensitive hashing (LSH).

cs.DS↗

arXiv2021

Freezing of a disorder induced quantum spin liquid

$\rm{Sr_2CuTe_{0.5}W_{0.5}O_6}$ is a square-lattice magnet with super-exchange between S=1/2 $\rm{Cu^{2+}}$ spins mediated by randomly distributed Te and W ions. Here, using sub-K temperature and 20 $\rmμ$eV energy resolution neutron scattering experiments we show that this system transits from a gapless disorder-induced quantum spin liquid to a new quantum state below $\rm{T_f}$ = 1.7(1) K, exhibiting a weak frozen moment of /S ~ 0.1 and low energy dynamic susceptibility,$χ''({\hbarω})$ linear in energy which is surprising for such a weak freezing in this highly fluctuating quantum regime.

cond-mat.mtrl-sci↗

arXiv2020

Aggregated Deletion Propagation for Counting Conjunctive Query Answers

We investigate the computational complexity of minimizing the source side-effect in order to remove a given number of tuples from the output of a conjunctive query. This is a variant of the well-studied {\em deletion propagation} problem, the difference being that we are interested in removing the smallest subset of input tuples to remove a given number of output tuples} while deletion propagation focuses on removing a specific output tuple. We call this the {\em Aggregated Deletion Propagation} problem. We completely characterize the poly-time solvability of this problem for arbitrary conjunctive queries without self-joins. This includes a poly-time algorithm to decide solvability, as well as an exact structural characterization of NP-hard instances. We also provide a practical algorithm for this problem (a heuristic for NP-hard instances) and evaluate its experimental performance on real and synthetic datasets.

cs.DB↗

arXiv2020

Algorithms for a Topology-aware Massively Parallel Computation Model

Most of the prior work in massively parallel data processing assumes homogeneity, i.e., every computing unit has the same computational capability, and can communicate with every other unit with the same latency and bandwidth. However, this strong assumption of a uniform topology rarely holds in practical settings, where computing units are connected through complex networks. To address this issue, Blanas et al. recently proposed a topology-aware massively parallel computation model that integrates the network structure and heterogeneity in the modeling cost. The network is modeled as a directed graph, where each edge is associated with a cost function that depends on the data transferred between the two endpoints. The computation proceeds in synchronous rounds, and the cost of each round is measured as the maximum cost over all the edges in the network. In this work, we take the first step into investigating three fundamental data processing tasks in this topology-aware parallel model: set intersection, cartesian product, and sorting. We focus on network topologies that are tree topologies, and present both lower bounds, as well as (asymptotically) matching upper bounds. The optimality of our algorithms is with respect to the initial data distribution among the network nodes, instead of assuming worst-case distribution as in previous results. Apart from the theoretical optimality of our results, our protocols are simple, use a constant number of rounds, and we believe can be implemented in practical settings as well.

cs.DB↗

arXiv2020

Comment on "Spin-Momentum-Locked Edge Mode for Topological Vortex Lasing, Phys. Rev. Lett. vol. 125, 013903 (2020)"

We present a comment on "Spin-Momentum-Locked Edge Mode for Topological Vortex Lasing, Phys. Rev. Lett. vol. 125, 013903 (2020)"(hereafter the Letter).In the Letter, Yang et al. reported on an elegant topological vortex laser and proposed that the near-field spin and OAM of the topological edge mode lasing have a one-to-one far-field radiation correspondence. The near-field information is based on frequency dispersions of the topological edge modes, without supporting measurements and/or computer simulations. Unfortunately, their frequency dispersions shown in Fig. 1(c) (see also Fig. S6 and Eqs. (5.3) and (5.4) in Supplemental Material) are wrong. As the result, the mode assignment of the main mode |-2,+> investigated in the Letter is mistaken, which should be |2,+>. This spoils the one-to-one correspondence claimed in the Letter.

physics.optics↗

arXiv2020

Temporally-decoherent and spatially-coherent vibrations in metal halide perovskite

The long carrier lifetime and defect tolerance in metal halide perovskites (MHPs) are major contributors to the superb performance of MHP optoelectronic devices. Large polarons were reported to be responsible for the long carrier lifetime. Yet microscopic mechanisms of the large polaron formation including the so-called phonon melting, are still under debate. Here, time-of-flight (TOF) inelastic neutron scattering (INS) experiments and first-principles density-functional theory (DFT) calculations were employed to investigate the lattice vibrations (or phonon dynamics) in methylammonium lead iodide ($\rm{MAPbI_3}$), a prototypical example of MHPs. Our findings are that optical phonons lose temporal coherence gradually with increasing temperature which vanishes at the orthorhombic-to-tetragonal structural phase transition. Surprisingly, however, we found that the spatial coherence is still retained throughout the decoherence process. We argue that the temporally decoherent and spatially coherent vibrations contribute to the formation of large polarons in this metal halide perovskite.

cond-mat.mtrl-sci↗

arXiv2020

Observing Responses to the COVID-19 Pandemic using Worldwide Network Cameras

COVID-19 has resulted in a worldwide pandemic, leading to "lockdown" policies and social distancing. The pandemic has profoundly changed the world. Traditional methods for observing these historical events are difficult because sending reporters to areas with many infected people can put the reporters' lives in danger. New technologies are needed for safely observing responses to these policies. This paper reports using thousands of network cameras deployed worldwide for the purpose of witnessing activities in response to the policies. The network cameras can continuously provide real-time visual data (image and video) without human efforts. Thus, network cameras can be utilized to observe activities without risking the lives of reporters. This paper describes a project that uses network cameras to observe responses to governments' policies during the COVID-19 pandemic (March to April in 2020). The project discovers over 30,000 network cameras deployed in 110 countries. A set of computer tools are created to collect visual data from network cameras continuously during the pandemic. This paper describes the methods to discover network cameras on the Internet, the methods to collect and manage data, and preliminary results of data analysis. This project can be the foundation for observing the possible "second wave" in fall 2020. The data may be used for post-pandemic analysis by sociologists, public health experts, and meteorologists.

cs.CY↗