arXiv subjects

Xu Huang

Publications and source records attributed to Xu Huang.

At least 91 records · Page 5Linked to original sources

arXiv2022

Fast Variational AutoEncoder with Inverted Multi-Index for Collaborative Filtering

Variational AutoEncoder (VAE) has been extended as a representative nonlinear method for collaborative filtering. However, the bottleneck of VAE lies in the softmax computation over all items, such that it takes linear costs in the number of items to compute the loss and gradient for optimization. This hinders the practical use due to millions of items in real-world scenarios. Importance sampling is an effective approximation method, based on which the sampled softmax has been derived. However, existing methods usually exploit the uniform or popularity sampler as proposal distributions, leading to a large bias of gradient estimation. To this end, we propose to decompose the inner-product-based softmax probability based on the inverted multi-index, leading to sublinear-time and highly accurate sampling. Based on the proposed proposals, we develop a fast Variational AutoEncoder (FastVAE) for collaborative filtering. FastVAE can outperform the state-of-the-art baselines in terms of both sampling quality and efficiency according to the experiments on three real-world datasets.

cs.LG↗

arXiv2022

LiDAR-guided Stereo Matching with a Spatial Consistency Constraint

The complementary fusion of light detection and ranging (LiDAR) data and image data is a promising but challenging task for generating high-precision and high-density point clouds. This study proposes an innovative LiDAR-guided stereo matching approach called LiDAR-guided stereo matching (LGSM), which considers the spatial consistency represented by continuous disparity or depth changes in the homogeneous region of an image. The LGSM first detects the homogeneous pixels of each LiDAR projection point based on their color or intensity similarity. Next, we propose a riverbed enhancement function to optimize the cost volume of the LiDAR projection points and their homogeneous pixels to improve the matching robustness. Our formulation expands the constraint scopes of sparse LiDAR projection points with the guidance of image information to optimize the cost volume of pixels as much as possible. We applied LGSM to semi-global matching and AD-Census on both simulated and real datasets. When the percentage of LiDAR points in the simulated datasets was 0.16%, the matching accuracy of our method achieved a subpixel level, while that of the original stereo matching algorithm was 3.4 pixels. The experimental results show that LGSM is suitable for indoor, street, aerial, and satellite image datasets and provides good transferability across semi-global matching and AD-Census. Furthermore, the qualitative and quantitative evaluations demonstrate that LGSM is superior to two state-of-the-art optimizing cost volume methods, especially in reducing mismatches in difficult matching areas and refining the boundaries of objects.

cs.CV↗

arXiv2022

Locally Cartesian Closed Categories

This note explains how dependent sums and products are interpreted by adjoints of the base change functor in a locally cartesian closed category. An effort is made to unpack all the definitions so as to make the concepts more transparent to new learners.

math.CT↗

arXiv2022

Incoherent coherences

This article explores a generic framework of well-typed and well-scoped syntaxes, with a signature-axiom approach resembling traditional abstract algebra. The boilerplate code needed in defining operations on syntaxes is identified and abstracted away. Some of the frequent boilerplate proofs are also generalized.

cs.LO↗

arXiv2021

Individual Tree Detection and Crown Delineation with 3D Information from Multi-view Satellite Images

Individual tree detection and crown delineation (ITDD) are critical in forest inventory management and remote sensing based forest surveys are largely carried out through satellite images. However, most of these surveys only use 2D spectral information which normally has not enough clues for ITDD. To fully explore the satellite images, we propose a ITDD method using the orthophoto and digital surface model (DSM) derived from the multi-view satellite data. Our algorithm utilizes the top-hat morphological operation to efficiently extract the local maxima from DSM as treetops, and then feed them to a modi-fied superpixel segmentation that combines both 2D and 3D information for tree crown delineation. In subsequent steps, our method incorporates the biological characteristics of the crowns through plant allometric equation to falsify potential outliers. Experiments against manually marked tree plots on three representative regions have demonstrated promising results - the best overall detection accuracy can be 89%.

cs.CV↗

arXiv2021

A Unified Framework of Bundle Adjustment and Feature Matching for High-Resolution Satellite Images

Bundle adjustment (BA) is a technique for refining sensor orientations of satellite images, while adjustment accuracy is correlated with feature matching results. Feature match-ing often contains high uncertainties in weak/repeat textures, while BA results are helpful in reducing these uncertainties. To compute more accurate orientations, this article incorpo-rates BA and feature matching in a unified framework and formulates the union as the optimization of a global energy function so that the solutions of the BA and feature matching are constrained with each other. To avoid a degeneracy in the optimization, we propose a comprised solution by breaking the optimization of the global energy function into two-step suboptimizations and compute the local minimums of each suboptimization in an incremental manner. Experiments on multi-view high-resolution satellite images show that our proposed method outperforms state-of-the-art orientation techniques with or without accurate least-squares matching.

cs.CV↗

arXiv2021

Geometric Processing for Image-based 3D Object Modeling

Image-based 3D object modeling refers to the process of converting raw optical images to 3D digital representations of the objects. Very often, such models are desired to be dimensionally true, semantically labeled with photorealistic appearance (reality-based modeling). Laser scanning was deemed as the standard (and direct) way to obtaining highly accurate 3D measurements of objects, while one would have to abide the high acquisition cost and its unavailability on some of the platforms. Nowadays the image-based methods backboned by the recently developed advanced dense image matching algorithms and geo-referencing paradigms, are becoming the dominant approaches, due to its high flexibility, availability and low cost. The largely automated geometric processing of images in a 3D object reconstruction workflow, from ordered/unordered raw imagery to textured meshes, is becoming a critical part of the reality-based 3D modeling. This article summarizes the overall geometric processing workflow, with focuses on introducing the state-of-the-art methods of three major components of geometric processing: 1) geo-referencing; 2) Image dense matching 3) texture mapping. Finally, we will draw conclusions and share our outlooks of the topics discussed in this article.

cs.CV↗

arXiv2020

CamVox: A Low-cost and Accurate Lidar-assisted Visual SLAM System

Combining lidar in camera-based simultaneous localization and mapping (SLAM) is an effective method in improving overall accuracy, especially at a large scale outdoor scenario. Recent development of low-cost lidars (e.g. Livox lidar) enable us to explore such SLAM systems with lower budget and higher performance. In this paper we propose CamVox by adapting Livox lidars into visual SLAM (ORB-SLAM2) by exploring the lidars' unique features. Based on the non-repeating nature of Livox lidars, we propose an automatic lidar-camera calibration method that will work in uncontrolled scenes. The long depth detection range also benefit a more efficient mapping. Comparison of CamVox with visual SLAM (VINS-mono) and lidar SLAM (LOAM) are evaluated on the same dataset to demonstrate the performance. We open sourced our hardware, code and dataset on GitHub.

cs.RO↗

arXiv2019

A Comparison of Stereo-Matching Cost between Convolutional Neural Network and Census for Satellite Images

Stereo dense image matching can be categorized to low-level feature based matching and deep feature based matching according to their matching cost metrics. Census has been proofed to be one of the most efficient low-level feature based matching methods, while fast Convolutional Neural Network (fst-CNN), as a deep feature based method, has small computing time and is robust for satellite images. Thus, a comparison between fst-CNN and census is critical for further studies in stereo dense image matching. This paper used cost function of fst-CNN and census to do stereo matching, then utilized semi-global matching method to obtain optimized disparity images. Those images are used to produce digital surface model to compare with ground truth points. It addresses that fstCNN performs better than census in the aspect of absolute matching accuracy, histogram of error distribution and matching completeness, but these two algorithms still performs in the same order of magnitude.

cs.CV↗

arXiv2019

Using Orthophoto for Building Boundary Sharpening in the Digital Surface Model

Nowadays dense stereo matching has become one of the dominant tools in 3D reconstruction of urban regions for its low cost and high flexibility in generating dense 3D points. However, state-of-the-art stereo matching algorithms usually apply a semi-global matching (SGM) strategy. This strategy normally assumes the surface geometry pieceswise planar, where a smooth penalty is imposed to deal with non-texture or repeating-texture areas. This on one hand, generates much smooth surface models, while on the other hand, may partially leads to smoothing on depth discontinuities, particularly for fence-shaped regions or densely built areas with narrow streets. To solve this problem, in this work, we propose to use the line segment information extracted from the corresponding orthophoto as a pose-processing tool to sharpen the building boundary of the Digital Surface Model (DSM) generated by SGM. Two methods which are based on graph-cut and plane fitting are proposed and compared. Experimental results on several satellite datasets with ground truth show the robustness and effectiveness of the proposed DSM sharpening method.

cs.CV↗

arXiv2019

Multi-View Large-Scale Bundle Adjustment Method for High-Resolution Satellite Images

Given enough multi-view image corresponding points (also called tie points) and ground control points (GCP), bundle adjustment for high-resolution satellite images is used to refine the orientations or most often used geometric parameters Rational Polynomial Coefficients (RPC) of each satellite image in a unified geodetic framework, which is very critical in many photogrammetry and computer vision applications. However, the growing number of high resolution spaceborne optical sensors has brought two challenges to the bundle adjustment: 1) images come from different satellite cameras may have different imaging dates, viewing angles, resolutions, etc., thus resulting in geometric and radiometric distortions in the bundle adjustment; 2) The large-scale mapping area always corresponds to vast number of bundle adjustment corrections (including RPC bias and object space point coordinates). Due to the limitation of computer memory, it is hard to refine all corrections at the same time. Hence, how to efficiently realize the bundle adjustment in large-scale regions is very important. This paper particularly addresses the multi-view large-scale bundle adjustment problem by two steps: 1) to get robust tie points among different satellite images, we design a multi-view, multi-source tie point matching algorithm based on plane rectification and epipolar constraints, which is able to compensate geometric and local nonlinear radiometric distortions among satellite datasets, and 2) to solve dozens of thousands or even millions of variables bundle adjustment corrections in the large scale bundle adjustment, we use an efficient solution with only a little computer memory. Experiments on in-track and off-track satellite datasets show that the proposed method is capable of computing sub-pixel accuracy bundle adjustment results.

cs.CV↗

arXiv2018

The orbital eccentricity of small planet systems

We determine the orbital eccentricities of individual small Kepler planets, through a combination of asteroseismology and transit light-curve analysis. We are able to constrain the eccentricities of 51 systems with a single transiting planet, which supplement our previous measurements of 66 planets in multi-planet systems. Through a Bayesian hierarchical analysis, we find evidence that systems with only one detected transiting planet have a different eccentricity distribution than systems with multiple detected transiting planets. The eccentricity distribution of the single-transiting systems is well described by the positive half of a zero-mean Gaussian distribution with a dispersion $σ_e = 0.32 \pm 0.06$, while the multiple-transit systems are consistent with $σ_e = 0.083^{+0.015}_{-0.020}$. A mixture model suggests a fraction of $0.76^{+0.21}_{-0.12}$ of single-transiting systems have a moderate eccentricity, represented by a Rayleigh distribution that peaks at $0.26^{+0.04}_{-0.06}$. This finding may reflect differences in the formation pathways of systems with different numbers of transiting planets. We investigate the possibility that eccentricities are "self-excited" in closely packed planetary systems, as well as the influence of long-period giant companion planets. We find that both mechanisms can qualitatively explain the observations. We do not find any evidence for a correlation between eccentricity and stellar metallicity, as has been seen for giant planets. Neither do we find any evidence that orbital eccentricity is linked to the detection of a companion star. Along with this paper we make available all of the parameters and uncertainties in the eccentricity distributions, as well as the properties of individual systems, for use in future studies.

astro-ph.EP↗

arXiv2015

Analysis of Pain Hemodynamic Response Using Near-Infrared Spectroscopy (NIRS)

Despite recent advances in brain research, understanding the various signals for pain and pain intensities in the brain cortex is still a complex task due to temporal and spatial variations of brain hemodynamics. In this paper we have investigated pain based on cerebral hemodynamics via near-infrared spectroscopy (NIRS). This study presents a pain stimulation experiment that uses three acupuncture manipulation techniques to safely induce pain in healthy subjects. Acupuncture pain response was presented and hemodynamic pain signal analysis showed the presence of dominant channels and their relationship among surrounding channels, which contribute the further pain research area.

q-bio.NC↗

arXiv2014

On the Occurrence Rate of Hot Jupiters in Different Stellar Environments

Many Hot Jupiters (HJs) are detected by the Doppler and the transit techniques. From surveys using these two techniques, however, the measured HJ occurrence rates differ by a factor of two or more. Using the California Planet Survey sample and the Kepler sample, we investigate the causes for the difference of HJ occurrence rate. First, we find that $12.8\%\pm0.24\%$ of HJs are misidentified in the Kepler mission because of photometric dilution and subgiant contamination. Second, we explore the differences between the Doppler sample and the Kepler sample that can account for the different HJ occurrence rate. Third, we discuss how to measure the fundamental HJ occurrence rates by synthesizing the results from the Doppler and Kepler surveys. The fundamental HJ occurrence rates are a measure of HJ occurrence rate as a function of stellar multiplicity and evolutionary stage, e.g., the HJ occurrence rate for single and multiple stars or for main sequence and subgiant stars. While we find qualitative evidence that HJs occur less frequently in subgiants and multiple stellar systems, we conclude that our current knowledge of stellar properties and stellar multiplicity rate is too limited for us to reach any quantitative result for the fundamental HJ occurrence rates. This concern extends to $η_{\rm{Earth}}$, the occurrence rate of Earth-like planets.

astro-ph.EP↗

arXiv2013

DFT-GGA errors in NO chemisorption energies on (111) transition metal surfaces: Possible origins and correction schemes

Here we investigate whether well-known DFT-GGA errors in predicting the chemisorption energy ($E_{\rm chem}$) of CO on transition metal surfaces manifest in analogous NO chemisorption systems. To verify the occurrence of DFT-GGA overestimation of the back-donation mechanism in NO Chemisorption, we use electronic structure analysis to show that the partially filled molecular NO 2$π^{*}$ orbital rehybridizes with the transition metal $d$-band to form new bonding and anti-bonding states. We relate the back-donation charge transfer associated with chemisorption to the promotion of an electron from the 5$σ$ orbital to the 2$π^{*}$ orbital in the gas-phase NO G$^{2}Σ^{-}\leftarrow \rm{X}^{2}Π$ excitation. We establish linear relationships between $E_{\rm chem}$ and $ΔE_{\rm G\leftarrow X}$ and go on to formulate an $E_{\rm chem}$ correction scheme in the style of Mason {\it et al.}, {[Physical Review B {\bf 69}, 161401(R)]}. We apply the NO $E_{\rm chem}$ correction method to the (111) surfaces of Pt, Pd, Rh, and Ir, with NO chemisorption modeled at a coverage of 0.25 ML. We note that both the slope of $E_{\rm chem}$ {\it vs.} $ΔE_{\rm G\leftarrow X}$ and the dipole moment depend strongly on adsorption site for each metal, and we use this fact to construct an approximate correction scheme which we go on to test using NO/Pt(100) chemisorption.

physics.chem-ph↗

arXiv2013

On the Coagulation and Size Distribution of Pressure Confined Cores

Observations of the Pipe Nebula have led to the discovery of dense starless cores. The mass of most cores is too small for their self gravity to hold them together. Instead, they are thought to be pressure confined. The observed dense cores' mass function (CMF) matches well with the initial mass function (IMF) of stars in young clusters. Similar CMF's are observed in other star forming regions such as the Aquila Nebula, albeit with some dispersion. The shape of these CMF provides important clues to the competing physical processes which lead to star formation and its feedback on the interstellar media. In this paper, we investigate the dynamical origin of the the mass function of starless cores which are confined by a warm, less dense medium. We consider the coagulation between the cold cores and their ablation due to Kelvin-Helmholtz instability induced by their relative motion through the warm medium. We are able to reproduce the observed CMF among the starless cores in the Pipe nebula. Our results indicate that in environment similar to the Pipe nebula: 1) before the onset of their gravitational collapse, the mass distribution of the progenitor cores is similar to that of the young stars, 2) the observed CMF is a robust consequence of dynamical equilibrium between the coagulation and ablation of cores, and 3) a break in the slope of the CMF is due to the enhancement of collisional cross section and suppression of ablation for cores with masses larger than the cores' Bonnor-Ebert mass.

astro-ph.GA↗

arXiv2012

150 New transiting planet candidates from Kepler Q1-Q6 data

We have performed an extensive search for planet candidates in the publicly available Kepler Long Cadence data from quarters Q1 through Q6. The search method consists of initial de-trending of the data, applying the trend filtering algorithm, searching for transit signals with the Box Least Squares fitting method in three frequency domains, visual inspection of the potential transit candidates, and in-depth analysis of the shortlisted candidates. In this paper we present 150 new periodic planet candidates and 7 single transit events, 72 of which are in multiple systems. The periods of these planet candidates vary from $\sim$0.17\,day to $\sim$ 440\,day. 124 of the planet candidates have radii smaller than 3 \rearth. We recover 82.5% of the Batalha et al. (2012) KOI catalog. We also report 40 newly identified false positives---systems that look like transiting planets, but are probably due to blended eclipsing binaries. Our search improves the statistics in the short period and small planet radii parameter ranges.

astro-ph.EP↗

arXiv2012

Ohmic Dissipation in the Interiors of Hot Jupiters

We present models of ohmic heating in the interiors of hot jupiters in which we decouple the interior and the wind zone by replacing the wind zone with a boundary temperature Tiso and magnetic field Bphi0. Ohmic heating influences the contraction of gas giants in two ways: by direct heating within the convection zone, and by heating outside the convection zone which increases the effective insulation of the interior. We calculate these effects, and show that internal ohmic heating is only able to slow the contraction rate of a cooling gas giant once the planet reaches a critical value of internal entropy. We determine the age of the gas giant when ohmic heating becomes important as a function of mass, Tiso and induced Bphi0. With this survey of parameter space complete, we then adopt the wind zone scalings of Menou (2012) and calculate the expected evolution of gas giants with different levels of irradiation. We find that,with this prescription of magnetic drag, it is difficult to inflate massive planets or those with strong irradiation using ohmic heating, meaning that we are unable to account for many of the observed hot jupiter radii. This is in contrast to previous evolutionary models that assumed that a constant fraction of the irradiation is transformed into ohmic power.

astro-ph.EP↗