arXiv subjects

Xin Ma

Publications and source records attributed to Xin Ma.

At least 127 records · Page 7Linked to original sources

arXiv2021

Point-line-based RGB-D SLAM and Bundle Adjustment Uncertainty Analysis

Most of the state-of-the-art indirect visual SLAM methods are based on the sparse point features. However, it is hard to find enough reliable point features for state estimation in the case of low-textured scenes. Line features are abundant in urban and indoor scenes. Recent studies have shown that the combination of point and line features can provide better accuracy despite the decrease in computational efficiency. In this paper, measurements of point and line features are extracted from RGB-D data to create map features, and points on a line are treated as keypoints. We propose an extended approach to make more use of line observation information. And we prove that, in the local bundle adjustment, the estimation uncertainty of keyframe poses can be reduced when considering more landmarks with independent measurements in the optimization process. Experimental results on two public RGB-D datasets demonstrate that the proposed method has better robustness and accuracy in challenging environments.

cs.RO↗

arXiv2021

Adaptive Deconvolution-based stereo matching Net for Local Stereo Matching

In deep learning-based local stereo matching methods, larger image patches usually bring better stereo matching accuracy. However, it is unrealistic to increase the size of the image patch size without restriction. Arbitrarily extending the patch size will change the local stereo matching method into the global stereo matching method, and the matching accuracy will be saturated. We simplified the existing Siamese convolutional network by reducing the number of network parameters and propose an efficient CNN based structure, namely Adaptive Deconvolution-based disparity matching Net (ADSM net) by adding deconvolution layers to learn how to enlarge the size of input feature map for the following convolution layers. Experimental results on the KITTI 2012 and 2015 datasets demonstrate that the proposed method can achieve a good trade-off between accuracy and complexity.

cs.CV↗

arXiv2020

Discovering key topics from short, real-world medical inquiries via natural language processing and unsupervised learning

Millions of unsolicited medical inquiries are received by pharmaceutical companies every year. It has been hypothesized that these inquiries represent a treasure trove of information, potentially giving insight into matters regarding medicinal products and the associated medical treatments. However, due to the large volume and specialized nature of the inquiries, it is difficult to perform timely, recurrent, and comprehensive analyses. Here, we propose a machine learning approach based on natural language processing and unsupervised learning to automatically discover key topics in real-world medical inquiries from customers. This approach does not require ontologies nor annotations. The discovered topics are meaningful and medically relevant, as judged by medical information specialists, thus demonstrating that unsolicited medical inquiries are a source of valuable customer insights. Our work paves the way for the machine-learning-driven analysis of medical inquiries in the pharmaceutical industry, which ultimately aims at improving patient care.

cs.LG↗

arXiv2020

Support $τ$-tilting modules and recollements

Let $(\mbox{mod} Λ',\mbox{mod} Λ,\mbox{mod} Λ'')$ be a recollement of abelian categories for artin algebras $Λ'$, $Λ$ and $Λ''$. Under certain conditions, we present an explicit construction of gluing of (support) $τ$-tilting modules in $\mbox{mod} Λ$ with respect to (support) $τ$-tilting modules in $\mbox{mod} Λ'$ and $\mbox{mod} Λ''$ respectively; conversely, we study the construction of (support) $τ$-tilting modules in $\mbox{mod} Λ'$ and $\mbox{mod} Λ''$ obtained from (support) $τ$-tilting modules in $\mbox{mod} Λ$.

math.CT↗

arXiv2020

Purely infinite locally compact Hausdorff étale groupoids and their $C^*$-algebras

In this paper, we introduce properties including groupoid comparison, pure infiniteness and paradoxical comparison as well as a new algebraic tool called groupoid semigroup for locally compact Hausdorff étale groupoids. We show these new tools help establishing pure infiniteness of reduced groupoid $C^*$-algebras. As an application, we show a dichotomy of stably finiteness against pure infiniteness for reduced groupoid $C^*$-algebras arising from locally compact Hausdorff étale minimal topological principal groupoids. This generalizes the dichotomy obtained by Bönicke-Li and Rainone-Sims. We also study the relation among our paradoxical comparison, $n$-filling property and locally contracting property appeared in the literature for locally compact Hausdorff étale groupoids.

math.OA↗

arXiv2020

Unsupervised Contrastive Photo-to-Caricature Translation based on Auto-distortion

Photo-to-caricature translation aims to synthesize the caricature as a rendered image exaggerating the features through sketching, pencil strokes, or other artistic drawings. Style rendering and geometry deformation are the most important aspects in photo-to-caricature translation task. To take both into consideration, we propose an unsupervised contrastive photo-to-caricature translation architecture. Considering the intuitive artifacts in the existing methods, we propose a contrastive style loss for style rendering to enforce the similarity between the style of rendered photo and the caricature, and simultaneously enhance its discrepancy to the photos. To obtain an exaggerating deformation in an unpaired/unsupervised fashion, we propose a Distortion Prediction Module (DPM) to predict a set of displacements vectors for each input image while fixing some controlling points, followed by the thin plate spline interpolation for warping. The model is trained on unpaired photo and caricature while can offer bidirectional synthesizing via inputting either a photo or a caricature. Extensive experiments demonstrate that the proposed model is effective to generate hand-drawn like caricatures compared with existing competitors.

cs.CV↗

arXiv2020

Free-Form Image Inpainting via Contrastive Attention Network

Most deep learning based image inpainting approaches adopt autoencoder or its variants to fill missing regions in images. Encoders are usually utilized to learn powerful representational spaces, which are important for dealing with sophisticated learning tasks. Specifically, in image inpainting tasks, masks with any shapes can appear anywhere in images (i.e., free-form masks) which form complex patterns. It is difficult for encoders to capture such powerful representations under this complex situation. To tackle this problem, we propose a self-supervised Siamese inference network to improve the robustness and generalization. It can encode contextual semantics from full resolution images and obtain more discriminative representations. we further propose a multi-scale decoder with a novel dual attention fusion module (DAF), which can combine both the restored and known regions in a smooth way. This multi-scale architecture is beneficial for decoding discriminative representations learned by encoders into images layer by layer. In this way, unknown regions will be filled naturally from outside to inside. Qualitative and quantitative experiments on multiple datasets, including facial and natural datasets (i.e., Celeb-HQ, Pairs Street View, Places2 and ImageNet), demonstrate that our proposed method outperforms state-of-the-art methods in generating high-quality inpainting results.

cs.CV↗

arXiv2020

Inspection-on-the-fly using Hybrid Physical Interaction Control for Aerial Manipulators

Inspection for structural properties (surface stiffness and coefficient of restitution) is crucial for understanding and performing aerial manipulations in unknown environments, with little to no prior knowledge on their state. Inspection-on-the-fly is the uncanny ability of humans to infer states during manipulation, reducing the necessity to perform inspection and manipulation separately. This paper presents an infrastructure for inspection-on-the-fly method for aerial manipulators using hybrid physical interaction control. With the proposed method, structural properties (surface stiffness and coefficient of restitution) can be estimated during physical interactions. A three-stage hybrid physical interaction control paradigm is presented to robustly approach, acquire and impart a desired force signature onto a surface. This is achieved by combining a hybrid force/motion controller with a model-based feed-forward impact control as intermediate phase. The proposed controller ensures a steady transition from unconstrained motion control to constrained force control, while reducing the lag associated with the force control phase. And an underlying Operational Space dynamic configuration manager permits complex, redundant vehicle/arm combinations. Experiments were carried out in a mock-up of a Dept. of Energy exhaust shaft, to show the effectiveness of the inspection-on-the-fly method to determine the structural properties of the target surface and the performance of the hybrid physical interaction controller in reducing the lag associated with force control phase.

cs.RO↗

arXiv2020

Exploiting Style and Attention in Real-World Super-Resolution

Real-world image super-resolution (SR) is a challenging image translation problem. Low-resolution (LR) images are often generated by various unknown transformations rather than by applying simple bilinear down-sampling on high-resolution (HR) images. To address this issue, this paper proposes a novel pipeline which exploits style and attention mechanism in real-world SR. Our pipeline consists of a style Variational Autoencoder (styleVAE) and a SR network incorporated with attention mechanism. To get real-world-like low-quality images paired with the HR images, we design the styleVAE to transfer the complex nuisance factors in real-world LR images to the generated LR images. We also use mutual information estimation (MI) to get better style information. For our SR network, we firstly propose a global attention residual block to learn long-range dependencies in images. Then another local attention residual block is proposed to enforce the attention of SR network moving to local areas of images in which texture detail will be filled. It is worth noticing that styleVAE can be presented in a plug-and-play manner and thus can help to improve the generalization and robustness of our SR method as well as other SR methods. Extensive experiments demonstrate that our method surpasses the state-of-the-art work, both quantitatively and qualitatively.

cs.CV↗

arXiv2020

Recollements and tilting modules

Let $(\mbox{mod} Λ',\mbox{mod} Λ,\mbox{mod} Λ'')$ be a recollement of module categories for artin algebras $Λ'$, $Λ$ and $Λ''$. We provide a sufficient condition such that a glued torsion pair in $\mbox{mod} Λ$ is tilting when the given two torsion pairs are tilting in $\mbox{mod} Λ'$ and $\mbox{mod} Λ''$ respectively. Using this result, we give a construction of gluing of tilting modules in $\mbox{mod} Λ$ with respect to tilting modules in $\mbox{mod} Λ'$ and $\mbox{mod} Λ''$ respectively.

math.RT↗

arXiv2020

A generalized type semigroup and dynamical comparison

In this paper, we construct a semigroup associated to an action of countable discrete group on a compact Hausdorff space, that can be regarded as a higher dimensional generalization of the type semigroup. Using this generalized type semigroup we obtain a new characterization of dynamical comparison. This answers a question of Kerr and Schafhauser. In addition, this paper suggests a definition of comparison for dynamical systems in which neither necessarily the acting group is amenable nor the action is minimal.

math.DS↗

arXiv2019

Fast and Incremental Loop Closure Detection Using Proximity Graphs

Visual loop closure detection, which can be considered as an image retrieval task, is an important problem in SLAM (Simultaneous Localization and Mapping) systems. The frequently used bag-of-words (BoW) models can achieve high precision and moderate recall. However, the requirement for lower time costs and fewer memory costs for mobile robot applications is not well satisfied. In this paper, we propose a novel loop closure detection framework titled `FILD' (Fast and Incremental Loop closure Detection), which focuses on an on-line and incremental graph vocabulary construction for fast loop closure detection. The global and local features of frames are extracted using the Convolutional Neural Networks (CNN) and SURF on the GPU, which guarantee extremely fast extraction speeds. The graph vocabulary construction is based on one type of proximity graph, named Hierarchical Navigable Small World (HNSW) graphs, which is modified to adapt to this specific application. In addition, this process is coupled with a novel strategy for real-time geometrical verification, which only keeps binary hash codes and significantly saves on memory usage. Extensive experiments on several publicly available datasets show that the proposed approach can achieve fairly good recall at 100\% precision compared to other state-of-the-art methods. The source code can be downloaded at https://github.com/AnshanTJU/FILD for further studies.

cs.RO↗

arXiv2019

Improved GM(1,1) model based on Simpson formula and its applications

The classical GM(1,1) model is an efficient tool to {make accurate forecasts} with limited samples. But the accuracy of the GM(1,1) model still needs to be improved. This paper proposes a novel discrete GM(1,1) model, named ${\rm GM_{SD}}$(1,1) model, of which the background value is reconstructed using Simpson formula. The expression of the specific time response function is deduced, and the relationship between our model} and the continuous GM(1,1) model with Simpson formula called ${\rm GM_{SC} }$(1,1) model is systematically discussed. The proposed model is proved to be unbiased to simulate the homogeneous exponent sequence. Further, some numerical examples are given to validate the accuracy of the new ${\rm GM_{SD}}$(1,1) model. Finally, this model is used to predict the Gross Domestic Product and the freightage of Lanzhou, and the results illustrate the ${\rm GM_{SD}}$(1,1) model provides accurate prediction.

stat.AP↗

arXiv2019

Application of a new information priority accumulated grey model with time power to predict short-term wind turbine capacity

Wind energy makes a significant contribution to global power generation. Predicting wind turbine capacity is becoming increasingly crucial for cleaner production. For this purpose, a new information priority accumulated grey model with time power is proposed to predict short-term wind turbine capacity. Firstly, the computational formulas for the time response sequence and the prediction values are deduced by grey modeling technique and the definite integral trapezoidal approximation formula. Secondly, an intelligent algorithm based on particle swarm optimization is applied to determine the optimal nonlinear parameters of the novel model. Thirdly, three real numerical examples are given to examine the accuracy of the new model by comparing with six existing prediction models. Finally, based on the wind turbine capacity from 2007 to 2017, the proposed model is established to predict the total wind turbine capacity in Europe, North America, Asia, and the world. The numerical results reveal that the novel model is superior to other forecasting models. It has a great advantage for small samples with new characteristic behaviors. Besides, reasonable suggestions are put forward from the standpoint of the practitioners and governments, which has high potential to advance the sustainable improvement of clean energy production in the future.

stat.AP↗

arXiv2019

Semi-parametric Bayes Regression with Network Valued Covariates

There is an increasing recognition of the role of brain networks as neuroimaging biomarkers in mental health and psychiatric studies. Our focus is posttraumatic stress disorder (PTSD), where the brain network interacts with environmental exposures in complex ways to drive the disease progression. Existing linear models seeking to characterize the relation between the clinical phenotype and the entire edge set in the brain network may be overly simplistic and often involve inflated number of parameters leading to computational burden and inaccurate estimation. In one of the first such efforts, we develop a novel two stage Bayesian framework to find a node-specific lower dimensional representation for the network using a latent scale approach in the first stage, and then use a flexible Gaussian process regression framework for prediction involving the latent scales and other supplementary covariates in the second stage. The proposed approach relaxes linearity assumptions, addresses the curse of dimensionality and is scalable to high dimensional networks while maintaining interpretability at the node level of the network. Extensive simulations and results from our motivating PTSD application show a distinct advantage of the proposed approach over competing linear and non-linear approaches in terms of prediction and coverage.

stat.ME↗

arXiv2019

The conformable fractional grey system model

The fractional order grey models (FGM) have appealed considerable interest of research in recent years due to its higher effectiveness and flexibility than the conventional grey models and other prediction models. However, the definitions of the fractional order accumulation (FOA) and difference (FOD) is computationally complex, which leads to difficulties for the theoretical analysis and applications. In this paper, the new definition of the FOA are proposed based on the definitions of Conformable Fractional Derivative, which is called the Conformable Fractional Accumulation (CFA), along with its inverse operation, the Conformable Fractional Difference (CFD). Then the new Conformable Fractional Grey Model (CFGM) based on CFA and CFD is introduced with detailed modelling procedures. The feasibility and simplicity and the CFGM are shown in the numerical example. And the at last the comprehensive real-world case studies of natural gas production forecasting in 11 countries are presented, and results show that the CFGM is much more effective than the existing FGM model in the 165 subcases.

stat.ME↗

arXiv2019

Comparison and pure infiniteness of crossed products

Let $α: G\curvearrowright X$ be a continuous action of an infinite countable group on a compact Hausdorff space. We show that, under the hypothesis that the action $α$ is topologically free and has no $G$-invariant regular Borel probability measure on $X$, dynamical comparison implies that the reduced crossed product of $α$ is purely infinite and simple. This result, as an application, shows a dichotomy between stable finiteness and pure infiniteness for reduced crossed products arising from actions satisfying dynamical comparison. We also introduce the concepts of paradoxical comparison and the uniform tower property. Under the hypothesis that the action $α$ is exact and essentially free, we show that paradoxical comparison together with the uniform tower property implies that the reduced crossed product of $α$ is purely infinite. As applications, we provide new results on pure infiniteness of reduced crossed products in which the underlying spaces are not necessarily zero-dimensional. Finally, we study the type semigroups of actions on the Cantor set in order to establish the equivalence of almost unperforation of the type semigroup and comparison. This sheds a light to a question arising in the paper of Rørdam and Sierakowski.

math.DS↗

arXiv2019

A note on the sensitivity of semiows

In this note, it is shown that there exist two non-syndetically sensitive cascades defined on complete metric spaces whose product is syndetically sensitive, answering negatively the Question 9.2 posed in [12, Miller, A., Money, C., Turk. J. Math., 41 (2017): 1323{1336]. Moreover, it is shown that there exists a syndetically sensitive semiflow (G;X) defined on a complete metric space X such that (G1;X) is not sensitive for some syndetic closed submonoid G1 of G, answering negatively the Open question 3 posed in [13, Money, C., PhD thesis, University of Louisville, 2015] and Question 43 posed in [8, Miller, A., Real Anal. Exchange, 42 (2017): 9{24].

math.DS↗