arXiv subjects

Yangyang Li

Publications and source records attributed to Yangyang Li.

72 records · Page 4Linked to original sources

arXiv2021

RSBNet: One-Shot Neural Architecture Search for A Backbone Network in Remote Sensing Image Recognition

Recently, a massive number of deep learning based approaches have been successfully applied to various remote sensing image (RSI) recognition tasks. However, most existing advances of deep learning methods in the RSI field heavily rely on the features extracted by the manually designed backbone network, which severely hinders the potential of deep learning models due the complexity of RSI and the limitation of prior knowledge. In this paper, we research a new design paradigm for the backbone architecture in RSI recognition tasks, including scene classification, land-cover classification and object detection. A novel one-shot architecture search framework based on weight-sharing strategy and evolutionary algorithm is proposed, called RSBNet, which consists of three stages: Firstly, a supernet constructed in a layer-wise search space is pretrained on a self-assembled large-scale RSI dataset based on an ensemble single-path training strategy. Next, the pre-trained supernet is equipped with different recognition heads through the switchable recognition module and respectively fine-tuned on the target dataset to obtain task-specific supernet. Finally, we search the optimal backbone architecture for different recognition tasks based on the evolutionary algorithm without any network training. Extensive experiments have been conducted on five benchmark datasets for different recognition tasks, the results show the effectiveness of the proposed search paradigm and demonstrate that the searched backbone is able to flexibly adapt different RSI recognition tasks and achieve impressive performance.

eess.IV↗

arXiv2021

Existence of Infinitely Many Minimal Hypersurfaces in Higher-dimensional Closed Manifolds with Generic Metrics

In this paper, we show that a closed manifold $M^{n+1} (n \geq 7)$ endowed with a $C^\infty$-generic (Baire sense) metric contains infinitely many singular minimal hypersurfaces with optimal regularity. Moreover, for $2 \leq n \leq 6$, our argument also implies the denseness of the minimal hypersurfaces realizing min-max widths for generic metrics. This partially supports equidistribution of the minimal hypersurfaces realizing min-max widths conjectured by F.C. Marques, A. Neves and A. Song.

math.DG↗

arXiv2021

Locate Who You Are: Matching Geo-location to Text for User Identity Linkage

Nowadays, users are encouraged to activate across multiple online social networks simultaneously. Anchor link prediction, which aims to reveal the correspondence among different accounts of the same user across networks, has been regarded as a fundamental problem for user profiling, marketing, cybersecurity, and recommendation. Existing methods mainly address the prediction problem by utilizing profile, content, or structural features of users in symmetric ways. However, encouraged by online services, users would also post asymmetric information across networks, such as geo-locations and texts. It leads to an emerged challenge in aligning users with asymmetric information across networks. Instead of similarity evaluation applied in previous works, we formalize correlation between geo-locations and texts and propose a novel anchor link prediction framework for matching users across networks. Moreover, our model can alleviate the label scarcity problem by introducing external data. Experimental results on real-world datasets show that our approach outperforms existing methods and achieves state-of-the-art results.

cs.SI↗

arXiv2021

Generic Regularity of Minimal Hypersurfaces in Dimension 8

In this paper, we show that every $8$-dimensional closed Riemmanian manifold with $C^\infty$-generic metrics admits a smooth minimal hypersurface. This generalized previous results by N. Smale and Chodosh-Liokumovich-Spolaor. Different from their local perturbation techniques, our construction is based on a global perturbation argument in and a novel geometric invariant which counts singular points with suitable weights.

math.DG↗

arXiv2021

Label Distribution Amendment with Emotional Semantic Correlations for Facial Expression Recognition

By utilizing label distribution learning, a probability distribution is assigned for a facial image to express a compound emotion, which effectively improves the problem of label uncertainties and noises occurred in one-hot labels. In practice, it is observed that correlations among emotions are inherently different, such as surprised and happy emotions are more possibly synchronized than surprised and neutral. It indicates the correlation may be crucial for obtaining a reliable label distribution. Based on this, we propose a new method that amends the label distribution of each facial image by leveraging correlations among expressions in the semantic space. Inspired by inherently diverse correlations among word2vecs, the topological information among facial expressions is firstly explored in the semantic space, and each image is embedded into the semantic space. Specially, a class-relation graph is constructed to transfer the semantic correlation among expressions into the task space. By comparing semantic and task class-relation graphs of each image, the confidence of its label distribution is evaluated. Based on the confidence, the label distribution is amended by enhancing samples with higher confidence and weakening samples with lower confidence. Experimental results demonstrate the proposed method is more effective than compared state-of-the-art methods.

cs.CV↗

arXiv2021

On the Existence of Minimal Hypersurfaces with Arbitrarily Large Area and Morse Index

We show that a bumpy closed Riemannian manifold $(M^{n+1}, g)$ $(3 \leq n+1 \leq 7)$ admits a sequence of connected closed embedded two-sided minimal hypersurfaces whose areas and Morse indices both tend to infinity. This improves a previous result by O. Chodosh and C. Mantoulidis on connected minimal hypersurfaces with arbitrarily large area.

math.DG↗

arXiv2021

Relevance-Aware Anomalous Users Detection in Social Network via Graph Neural Network

Anomalous users detection in social network is an imperative task for security problems. Motivated by the great power of Graph Neural Networks(GNNs), many current researches adopt GNN-based detectors to reveal the anomalous users. However, the increasing scale of social activities, explosive growth of users and manifold technical disguise render the user detection a difficult task. In this paper, we propose an innovate Relevance-aware Anomalous Users Detection model (RAU-GNN) to obtain a fine-grained detection result. RAU-GNN first extracts multiple relations of all types of users in social network, including both benign and anomalous users, and accordingly constructs the multiple user relation graph. Secondly, we employ relevance-aware GNN framework to learn the hidden features of users, and discriminate the anomalous users after discriminating. Concretely, by integrating Graph Convolution Network(GCN) and Graph Attention Network(GAT), we design a GCN-based relation fusion layer to aggregate initial information from different relations, and a GAT-based embedding layer to obtain the high-level embeddings. Lastly, we feed the learned representations to the following GNN layer in order to consolidate the node embedding by aggregating the final users' embeddings. We conduct extensive experiment on real-world datasets. The experimental results show that our approach can achieve high accuracy for anomalous users detection.

cs.SI↗

arXiv2021

Bipolar conduction and giant positive magnetoresistance in doped metallic titanium oxide heterostructures

Empowering conventional materials with unexpected magnetoelectric properties is appealing to the multi-functionalization of existing devices and the exploration of future electronics. Recently, owing to its unique effect in modulating a matter's properties, ultra-small dopants, e.g. H, D, and Li, attract enormous attention in creating emergent functionalities, such as superconductivity, and metal-insulator transition. Here, we report an observation of bipolar conduction accompanied by a giant positive magnetoresistance in D-doped metallic Ti oxide (TiOxDy) films. To overcome the challenges in intercalating the D into a crystalline oxide, a series of TiOxDy were formed by sequentially doping Ti with D and surface/interface oxidation. Intriguingly, while the electron mobility of the TiOxDy increases by an order of magnitude larger after doping, the emergent holes also exhibit high mobility. Moreover, the bipolar conduction induces a giant magnetoresistance up to 900% at 6 T, which is ~6 times higher than its conventional phase. Our study paves a way to empower conventional materials in existing electronics and induce novel electronic phases.

cond-mat.mtrl-sci↗

arXiv2021

An Anomaly Event Detection Method Based on GNN Algorithm for Multi-data Sources

Anomaly event detection is crucial for critical infrastructure security(transportation system, social-ecological sector, insurance service, government sector etc.) due to its ability to reveal and address the potential cyber-threats in advance by analysing the data(messages, microblogs, logs etc.) from digital systems and networks. However, the convenience and applicability of smart devices and the maturity of connected technology make the social anomaly events data multi-source and dynamic, which result in the inadaptability for multi-source data detection and thus affect the critical infrastructure security. To effectively address the proposed problems, in this paper, we design a novel anomaly detection method based on multi-source data. First, we leverage spectral clustering algorithm for feature extraction and fusion of multiple data sources. Second, by harnessing the power of deep graph neural network(Deep-GNN), we perform a fine-gained anomaly social event detection, revealing the threatening events and guarantee the critical infrastructure security. Experimental results demonstrate that our framework outperforms other baseline anomaly event detection methods and shows high tracking accuracy, strong robustness and stability.

cs.SI↗

arXiv2020

An Improved Morse Index Bound of Min-Max Minimal Hypersurfaces

In this paper, we give an improved Morse index bound of minimal hypersurfaces from Almgren-Pitts min-max construction in any closed Riemannian manifold $M^{n+1}$ $(n+1 \geq 3$), which generalizes a result by X. Zhou \cite{zhou_multiplicity_2019} for $3 \leq n+1 \leq 7$. The novel techniques are the construction of hierarchical deformations and the restrictive min-max theory. These techniques do not rely on bumpy metrics, and thus could be adapted to many other min-max settings.

math.DG↗

arXiv2020

A Study of Selectively Digital Etching Silicon-Germanium with Nitric and Hydrofluoric Acids

A digital etching method was proposed to achieve excellent control of etching depth. The digital etching characteristics of p+ Si and Si0.7Ge0.3 using the combinations of HNO3 oxidation and BOE oxide removal processes were studied. Experiments showed that oxidation saturates with time due to low activation energy. A physical model was presented to describe the wet oxidation process with nitric acid. The model was calibrated with experimental data and the oxidation saturation time, final oxide thickness, and selectivity between Si0.7Ge0.3 and p+ Si were obtained. The digital etch of laminated Si0.7Ge0.3/p+ Si was also investigated. The depth of the tunnels formed by etching SiGe layers between two Si layers was found in proportion to digital etching cycles. And oxidation would also saturate and the saturated relative etched amount per cycle (REPC) was 0.5 nm (4 monolayers). A corrected selectivity calculation formula was presented. The oxidation model was also calibrated with Si0.7Ge0.3/p+ Si stacks, and selectivity from model was the same with the corrected formula. The model can also be used to analyze process variations and repeatability. And it could act as a guidance for experiment design. Selectivity and repeatability should make a trade-off.

physics.app-ph↗

arXiv2017

Locality preserving projection on SPD matrix Lie group: algorithm and analysis

Symmetric positive definite (SPD) matrices used as feature descriptors in image recognition are usually high dimensional. Traditional manifold learning is only applicable for reducing the dimension of high-dimensional vector-form data. For high-dimensional SPD matrices, directly using manifold learning algorithms to reduce the dimension of matrix-form data is impossible. The SPD matrix must first be transformed into a long vector, and then the dimension of this vector must be reduced. However, this approach breaks the spatial structure of the SPD matrix space. To overcome this limitation, we propose a new dimension reduction algorithm on SPD matrix space to transform high-dimensional SPD matrices into low-dimensional SPD matrices. Our work is based on the fact that the set of all SPD matrices with the same size has a Lie group structure, and we aim to transform the manifold learning to the SPD matrix Lie group. We use the basic idea of the manifold learning algorithm called locality preserving projection (LPP) to construct the corresponding Laplacian matrix on the SPD matrix Lie group. Thus, we call our approach Lie-LPP to emphasize its Lie group character. We present a detailed algorithm analysis and show through experiments that Lie-LPP achieves effective results on human action recognition and human face recognition.

cs.CV↗

arXiv2017

Applying Ricci Flow to High Dimensional Manifold Learning

Traditional manifold learning algorithms often bear an assumption that the local neighborhood of any point on embedded manifold is roughly equal to the tangent space at that point without considering the curvature. The curvature indifferent way of manifold processing often makes traditional dimension reduction poorly neighborhood preserving. To overcome this drawback we propose a new algorithm called RF-ML to perform an operation on the manifold with help of Ricci flow before reducing the dimension of manifold.

cs.LG↗

arXiv2017

EEG and ECG changes during deep-sea manned submersible operation

Background: Deep-sea manned submersible operation could induce mental workload and influence neurophysiological measures. Psychophysiological responses to submersible operation are not well known. The main aim of this study was to investigate changes in EEG and ECG components and subjective mental stress of pilots during submersible operation. Methods: There were 6 experienced submersible pilots who performed a 3 h submersible operation task composed of 5 subtasks. Electroencephalogram (EEG) and electrocardiogram (ECG) was recorded before the operation task, after 1.5 h and 2.5 h operation, and after the task. Subjective ratings of mental stress were also conducted at these time points. Results: HR and scores on subjective stressed scale increased during the task compared to baseline (P<0.05). LF/HF ratio at 1.5 h were higher than those at Baseline (P<0.05) and 2.5 h (P<0.05). Relative theta power at the Cz site increased (P<0.01) and relative alpha power decreased (P<0.01) at 2.5 h compared to values at Baseline. Alpha attenuation coefficient (AAC, ratio of mean alpha power during eyes closed versus eyes open) at 2.5 h and after the task were lower compared to baseline and 1.5 h (P<0.05 or less). Conclusions: Submersible operation resulted in an increased HR in association with mental stress, alterations in autonomic activity and EEG changes that expressed variations in mental workload. Brain arousal level declined during the later operation period.

q-bio.NC↗

arXiv2017

Curvature-aware Manifold Learning

Traditional manifold learning algorithms assumed that the embedded manifold is globally or locally isometric to Euclidean space. Under this assumption, they divided manifold into a set of overlapping local patches which are locally isometric to linear subsets of Euclidean space. By analyzing the global or local isometry assumptions it can be shown that the learnt manifold is a flat manifold with zero Riemannian curvature tensor. In general, manifolds may not satisfy these hypotheses. One major limitation of traditional manifold learning is that it does not consider the curvature information of manifold. In order to remove these limitations, we present our curvature-aware manifold learning algorithm called CAML. The purpose of our algorithm is to break the local isometry assumption and to reduce the dimension of the general manifold which is not isometric to Euclidean space. Thus, our method adds the curvature information to the process of manifold learning. The experiments have shown that our method CAML is more stable than other manifold learning algorithms by comparing the neighborhood preserving ratios.

cs.LG↗

arXiv2017

Nitrogen-doped Nanoporous Carbon Membranes Functionalized with Co/CoP Janus-type nanocrystals as Hydrogen Evolution Electrode in Both Acid and Alkaline Environment

Self-supported electrocatalysts being generated and employed directly as electrode for energy conversion has been intensively pursued in the fields of materials chemistry and energy. Herein, we report a synthetic strategy to prepare freestanding hierarchically structured, nitrogen-doped nanoporous graphitic carbon membranes functionalized with Janus-type Co/CoP nanocrystals (termed as HNDCM-Co/CoP), which were successfully applied as a highly-efficient, binder-free electrode in hydrogen evolution reaction (HER). Benefited from multiple structural merits, such as high degree of graphitization, three-dimensionally interconnected micro-/meso-/macropores, uniform nitrogen-doping, well-dispersed Co/CoP nanocrystals as well as the confinement effect of the thin carbon layer on the nanocrystals, HNDCM-Co/CoP exhibited superior electrocatalytic activity and long-term operation stability for HER under both acid and alkaline conditions. As a proof-of-concept of practical usage, a macroscopic piece of HNDCM-Co/CoP of 5.6 cm x 4 cm x 60 um in size was prepared in our laboratory. Driven by a solar cell, electroreduction of water in alkaline condition (pH 14) was performed, and H2 has been produced at a rate of 16 ml/min, demonstrating its potential as real-life energy conversion systems.

physics.chem-ph↗

arXiv2017

Synthesis of Single-crystal-like Nanoporous Carbon Membranes and Their Applications in Overall Water Splitting

Nanoporous graphitic carbon membranes with defined chemical composition and pore architecture are novel nanomaterials that are actively pursued. Compared to easy-to-make porous carbon powders that dominate the porous carbon research and applications in energy generation/conversion and environmental remediation, porous carbon membranes are synthetically more challenging though rather appealing from an application perspective due to their structural integrity, interconnectivity and purity. Here we report a simple bottom-up approach to fabricate large-size, freestanding, porous carbon membranes that feature an unusual single-crystal-like graphitic order and hierarchical pore architecture plus favorable nitrogen doping. When loaded with cobalt nanoparticles, such carbon membranes serve as high-performance carbon-based non-noble metal electrocatalyst for overall water splitting.

cond-mat.mtrl-sci↗

arXiv2016

Achieving ultrahigh carrier mobility and photo-responsivity in solution-processed perovskite/carbon nanotubes phototransistors

Organolead trihalide perovskites have drawn substantial interest for applications in photovoltaic and optoelectronic devices due to their low processing cost and remarkable physical properties. However, perovskite thin films still suffer from low carrier mobility, limiting their device performance and application potential. Here we report that embedding single-walled carbon nanotubes into halide perovskite films can significantly enhance the hole and electron mobilities to record-high values of 595.3 and 108.7 cm2 V-1 s-1, respectively. In the ambipolar phototransistors with such hybrid channels, photo-carriers generated in the light-absorbing perovskite matrix are transported by the carbon nanotubes, leading to ultrahigh detectivity of 6 * 1014 Jones and responsivity of 1 * 104 A W-1. We find that the perovskite precursor in dimethylformamide solution serve as an excellent stabilizer for the dispersion of carbon nanotubes, which potentially extend the scope of applications of perovskites in solution-processed functional composites. The unprecedented high performances underscore the perovskite/carbon nanotubes hybrids as an emerging class of functional materials in optoelectronic and other applications.

cond-mat.mtrl-sci↗