arXiv subjects

Cheng Zhang

Publications and source records attributed to Cheng Zhang.

At least 361 records · Page 20Linked to original sources

arXiv2022

Optimal transport for causal discovery

To determine causal relationships between two variables, approaches based on Functional Causal Models (FCMs) have been proposed by properly restricting model classes; however, the performance is sensitive to the model assumptions, which makes it difficult to use. In this paper, we provide a novel dynamical-system view of FCMs and propose a new framework for identifying causal direction in the bivariate case. We first show the connection between FCMs and optimal transport, and then study optimal transport under the constraints of FCMs. Furthermore, by exploiting the dynamical interpretation of optimal transport under the FCM constraints, we determine the corresponding underlying dynamical process of the static cause-effect pair data. It provides a new dimension for describing static causal discovery tasks while enjoying more freedom for modeling the quantitative causal influences. In particular, we show that Additive Noise Models (ANMs) correspond to volume-preserving pressureless flows. Consequently, based on their velocity field divergence, we introduce a criterion for determining causal direction. With this criterion, we propose a novel optimal transport-based algorithm for ANMs which is robust to the choice of models and extend it to post-nonlinear models. Our method demonstrated state-of-the-art results on both synthetic and causal discovery benchmark datasets.

cs.LG↗

arXiv2022

Exploring and Evaluating Image Restoration Potential in Dynamic Scenes

In dynamic scenes, images often suffer from dynamic blur due to superposition of motions or low signal-noise ratio resulted from quick shutter speed when avoiding motions. Recovering sharp and clean results from the captured images heavily depends on the ability of restoration methods and the quality of the input. Although existing research on image restoration focuses on developing models for obtaining better restored results, fewer have studied to evaluate how and which input image leads to superior restored quality. In this paper, to better study an image's potential value that can be explored for restoration, we propose a novel concept, referring to image restoration potential (IRP). Specifically, We first establish a dynamic scene imaging dataset containing composite distortions and applied image restoration processes to validate the rationality of the existence to IRP. Based on this dataset, we investigate several properties of IRP and propose a novel deep model to accurately predict IRP values. By gradually distilling and selective fusing the degradation features, the proposed model shows its superiority in IRP prediction. Thanks to the proposed model, we are then able to validate how various image restoration related applications are benefited from IRP prediction. We show the potential usages of IRP as a filtering principle to select valuable frames, an auxiliary guidance to improve restoration models, and even an indicator to optimize camera settings for capturing better images under dynamic scenarios.

cs.CV↗

arXiv2022

Upsampling Autoencoder for Self-Supervised Point Cloud Learning

In computer-aided design (CAD) community, the point cloud data is pervasively applied in reverse engineering, where the point cloud analysis plays an important role. While a large number of supervised learning methods have been proposed to handle the unordered point clouds and demonstrated their remarkable success, their performance and applicability are limited to the costly data annotation. In this work, we propose a novel self-supervised pretraining model for point cloud learning without human annotations, which relies solely on upsampling operation to perform feature learning of point cloud in an effective manner. The key premise of our approach is that upsampling operation encourages the network to capture both high-level semantic information and low-level geometric information of the point cloud, thus the downstream tasks such as classification and segmentation will benefit from the pre-trained model. Specifically, our method first conducts the random subsampling from the input point cloud at a low proportion e.g., 12.5%. Then, we feed them into an encoder-decoder architecture, where an encoder is devised to operate only on the subsampled points, along with a upsampling decoder is adopted to reconstruct the original point cloud based on the learned features. Finally, we design a novel joint loss function which enforces the upsampled points to be similar with the original point cloud and uniformly distributed on the underlying shape surface. By adopting the pre-trained encoder weights as initialisation of models for downstream tasks, we find that our UAE outperforms previous state-of-the-art methods in shape classification, part segmentation and point cloud upsampling tasks. Code will be made publicly available upon acceptance.

cs.CV↗

arXiv2022

Electron-beam Introduction of Heteroatomic Pt-Si Structures in Graphene

Electron-beam (e-beam) manipulation of single dopant atoms in an aberration-corrected scanning transmission electron microscope is emerging as a method for directed atomic motion and atom-by-atom assembly. Until now, the dopant species have been limited to atoms closely matched to carbon in terms of ionic radius and capable of strong covalent bonding with carbon atoms in the graphene lattice. In situ dopant insertion into a graphene lattice has thus far been demonstrated only for Si, which is ubiquitously present as a contaminant in this material. Here, we achieve in situ manipulation of Pt atoms and their insertion into the graphene host matrix using the e-beam deposited Pt on graphene as a host system. We further demonstrate a mechanism for stabilization of the Pt atom, enabled through the formation of Si-stabilized Pt heteroatomic clusters attached to the graphene surface. This study provides evidence toward the universality of the e-beam assembly approach, opening a pathway for exploring cluster chemistry through direct assembly.

cond-mat.mtrl-sci↗

arXiv2022

Local Constraint-Based Causal Discovery under Selection Bias

We consider the problem of discovering causal relations from independence constraints selection bias in addition to confounding is present. While the seminal FCI algorithm is sound and complete in this setup, no criterion for the causal interpretation of its output under selection bias is presently known. We focus instead on local patterns of independence relations, where we find no sound method for only three variable that can include background knowledge. Y-Structure patterns are shown to be sound in predicting causal relations from data under selection bias, where cycles may be present. We introduce a finite-sample scoring rule for Y-Structures that is shown to successfully predict causal relations in simulation experiments that include selection mechanisms. On real-world microarray data, we show that a Y-Structure variant performs well across different datasets, potentially circumventing spurious correlations due to selection bias.

cs.LG↗

arXiv2022

Simultaneous Missing Value Imputation and Structure Learning with Groups

Learning structures between groups of variables from data with missing values is an important task in the real world, yet difficult to solve. One typical scenario is discovering the structure among topics in the education domain to identify learning pathways. Here, the observations are student performances for questions under each topic which contain missing values. However, most existing methods focus on learning structures between a few individual variables from the complete data. In this work, we propose VISL, a novel scalable structure learning approach that can simultaneously infer structures between groups of variables under missing data and perform missing value imputations with deep learning. Particularly, we propose a generative model with a structured latent space and a graph neural network-based architecture, scaling to a large number of variables. Empirically, we conduct extensive experiments on synthetic, semi-synthetic, and real-world education data sets. We show improved performances on both imputation and structure learning accuracy compared to popular and recent approaches.

cs.LG↗

arXiv2022

Composite pulses for high fidelity population transfer in three-level systems

In this work, we propose a composite pulses scheme by modulating phases to achieve high fidelity population transfer in three-level systems. To circumvent the obstacle that not enough variables are exploited to eliminate the systematic errors in the transition probability, we put forward a cost function to find the optimal value. The cost function is independently constructed either in ensuring an accurate population of the target state, or in suppressing the population of the leakage state, or both of them. The results demonstrate that population transfer is implemented with high fidelity even when existing the deviations in the coupling coefficients. Furthermore, our composite pulses scheme can be extensible to arbitrarily long pulse sequences. As an example, we employ the composite pulses sequence for achieving the three-atom singlet state in an atom-cavity system with ultrahigh fidelity. The final singlet state shows robustness against deviations and is not seriously affected by waveform distortions. Also, the singlet state maintains a high fidelity under the decoherence environment.

quant-ph↗

arXiv2022

Probing the gluon tomography in photoproduction of di-pions

A sizable $\cos 4ϕ$ azimuthal asymmetry in exclusive di-pion production near $ρ^0$ resonance peak in ultraperipheral heavy ion collisions recently has been reported by STAR collaboration. We show that both elliptic gluon Wigner distribution and final state soft photon radiation can give rise to this azimuthal asymmetry. The fact that the QED effect alone severely underestimates the observed asymmetry might signal the existence of the nontrivial correlation in quantum phase distribution of gluons.

hep-ph↗

arXiv2022

Reinforcement Learning in the Wild: Scalable RL Dispatching Algorithm Deployed in Ridehailing Marketplace

In this study, a real-time dispatching algorithm based on reinforcement learning is proposed and for the first time, is deployed in large scale. Current dispatching methods in ridehailing platforms are dominantly based on myopic or rule-based non-myopic approaches. Reinforcement learning enables dispatching policies that are informed of historical data and able to employ the learned information to optimize returns of expected future trajectories. Previous studies in this field yielded promising results, yet have left room for further improvements in terms of performance gain, self-dependency, transferability, and scalable deployment mechanisms. The present study proposes a standalone RL-based dispatching solution that is equipped with multiple mechanisms to ensure robust and efficient on-policy learning and inference while being adaptable for full-scale deployment. A new form of value updating based on temporal difference is proposed that is more adapted to the inherent uncertainty of the problem. For the driver-order assignment, a customized utility function is proposed that when tuned based on the statistics of the market, results in remarkable performance improvement and interpretability. In addition, for reducing the risk of cancellation after drivers' assignment, an adaptive graph pruning strategy based on the multi-arm bandit problem is introduced. The method is evaluated using offline simulation with real data and yields notable performance improvement. In addition, the algorithm is deployed online in multiple cities under DiDi's operation for A/B testing and is launched in one of the major international markets as the primary mode of dispatch. The deployed algorithm shows over 1.3% improvement in total driver income from A/B testing. In addition, by causal inference analysis, as much as 5.3% improvement in major performance metrics is detected after full-scale deployment.

cs.LG↗

arXiv2021

ThumbTrak: Recognizing Micro-finger Poses Using a Ring with Proximity Sensing

ThumbTrak is a novel wearable input device that recognizes 12 micro-finger poses in real-time. Poses are characterized by the thumb touching each of the 12 phalanges on the hand. It uses a thumb-ring, built with a flexible printed circuit board, which hosts nine proximity sensors. Each sensor measures the distance from the thumb to various parts of the palm or other fingers. ThumbTrak uses a support-vector-machine (SVM) model to classify finger poses based on distance measurements in real-time. A user study with ten participants showed that ThumbTrak could recognize 12 micro finger poses with an average accuracy of 93.6%. We also discuss potential opportunities and challenges in applying ThumbTrak in real-world applications.

cs.HC↗

arXiv2021

Classification of integrable boundary equations for integrable quad-graph systems

In the context of integrable systems on quad-graphs, the boundary consistency around a half of a rhombic dodecahedron, as a companion notion to the three-dimensional consistency around a cube, was introduced as a criterion for defining integrable boundary conditions for quad-graph systems with a boundary. In this paper, we formalize the notions of boundary equations as boundary conditions for quad-graph systems, and provide a systematic method for solving the boundary consistency, which results in a classification of integrable boundary equations for quad-graph equations in the Adler-Bobenko-Suris classification. This relies on factorizing, first the quad-graph equations into pairs of dual boundary equations, and then the consistency on a rhombic dodecahedron into two equivalent boundary consistencies. Generalizations of the method to rhombic-symmetric equations are also considered.

nlin.SI↗

arXiv2021

Theoretical study on hot-carrier generation, transport and injection in TiN/TiO$_2$ junction

Promoting performance of generation, transport and injection of hot carriers in metal/semiconductor junctions is critical for harvesting energy of hot carriers. However, the injection efficiency of hot carriers in the commonly used noble metals such as Au is extremely low, which hinders the hot carrier-based applications. Here, we predict that the plasmonic material TiN might be promising for generating, transporting and injecting hot carriers, based on first-principles calculations and Monte Carlo simulations. We found that the hot-carrier generation rate in TiN is significantly large compared to Au. The transport property of hot carriers in TiN is excellent, due to the long lifetime and mean free path. The TiN/TiO2 junction possesses very low Schottky barrier, which results in much larger injection efficiency than that in the Au/TiO2 junction. Furthermore, it is revealed that the optimal injection efficiency could be achieved in a core/shell cylindrical TiN/TiO2 junction. Our findings provide an in-depth understanding of hot-carrier generation, transport and injection, and hence are helpful for the development of hot carrier-based devices.

cond-mat.mtrl-sci↗

arXiv2021

On Model Calibration for Long-Tailed Object Detection and Instance Segmentation

Vanilla models for object detection and instance segmentation suffer from the heavy bias toward detecting frequent objects in the long-tailed setting. Existing methods address this issue mostly during training, e.g., by re-sampling or re-weighting. In this paper, we investigate a largely overlooked approach -- post-processing calibration of confidence scores. We propose NorCal, Normalized Calibration for long-tailed object detection and instance segmentation, a simple and straightforward recipe that reweighs the predicted scores of each class by its training sample size. We show that separately handling the background class and normalizing the scores over classes for each proposal are keys to achieving superior performance. On the LVIS dataset, NorCal can effectively improve nearly all the baseline models not only on rare classes but also on common and frequent classes. Finally, we conduct extensive analysis and ablation studies to offer insights into various modeling choices and mechanisms of our approach. Our code is publicly available at https://github.com/tydpan/NorCal/.

cs.CV↗

arXiv2021

Plastic Laminate Antireflective Coatings for Millimeter-wave Optics in BICEP Array

The BICEP/Keck series of experiments target the Cosmic Microwave Background at degree-scale resolution from the South Pole. Over the next few years, the "Stage-3" BICEP Array (BA) telescope will improve the program's frequency coverage and sensitivity to primordial B-mode polarization by an order of magnitude. The first receiver in the array, BA1, began observing at 30/40 GHz in early 2020. The next two receivers, BA2 and BA3, are currently being assembled and will map the southern sky at frequencies ranging from 95 GHz to 150 GHz. Common to all BA receivers is a refractive, on-axis, cryogenic optical design that focuses microwave radiation onto a focal plane populated with antenna-coupled bolometers. High-performance antireflective coatings up to 760 mm in aperture are needed for each element in the optical chain, and must withstand repeated thermal cycles down to 4 K. Here we present the design and fabrication of the 30/40 GHz anti-reflection coatings for the recently deployed BA1 receiver, then discuss laboratory measurements of their reflectance. We review the lamination method for these single- and dual-layer plastic coatings with indices matched to various polyethylene, nylon and alumina optics. We also describe ongoing efforts to optimize coatings for the next BA cryostats, which may inform technological choices for future Small-Aperture Telescopes of the CMB "Stage 4" experiment.

astro-ph.IM↗

arXiv2021

Sparse Uncertainty Representation in Deep Learning with Inducing Weights

Bayesian neural networks and deep ensembles represent two modern paradigms of uncertainty quantification in deep learning. Yet these approaches struggle to scale mainly due to memory inefficiency issues, since they require parameter storage several times higher than their deterministic counterparts. To address this, we augment the weight matrix of each layer with a small number of inducing weights, thereby projecting the uncertainty quantification into such low dimensional spaces. We further extend Matheron's conditional Gaussian sampling rule to enable fast weight sampling, which enables our inference method to maintain reasonable run-time as compared with ensembles. Importantly, our approach achieves competitive performance to the state-of-the-art in prediction and uncertainty estimation tasks with fully connected neural networks and ResNets, while reducing the parameter size to $\leq 24.3\%$ of that of a $single$ neural network.

stat.ML↗

arXiv2021

Smart Flood Resilience: Harnessing Community-Scale Big Data for Predictive Flood Risk Monitoring, Rapid Impact Assessment, and Situational Awareness

Smart resilience is the beneficial result of the collision course of the fields of data science and urban resilience to flooding. The objective of this study is to propose and demonstrate a smart flood resilience framework that leverages heterogeneous community-scale big data and infrastructure sensor data to enhance predictive risk monitoring and situational awareness. The smart flood resilience framework focuses on four core capabilities that could be augmented by the use of heterogeneous community-scale big data and analytics techniques: (1) predictive flood risk mapping; (2) automated rapid impact assessment; (3) predictive infrastructure failure prediction and monitoring; and (4) smart situational awareness capabilities. We demonstrate the components of these core capabilities of the smart flood resilience framework in the context of the 2017 Hurricane Harvey in Harris County, Texas. First, we demonstrate the use of flood sensors for the prediction of floodwater overflow in channel networks and inundation of co-located road networks. Second, we discuss the use of social media and machine learning techniques for assessing the impacts of floods on communities and sensing emotion signals to examine societal impacts. Third, we illustrate the use of high-resolution traffic data in network-theoretic models for nowcasting of flood propagation on road networks and the disrupted access to critical facilities, such as hospitals. Fourth, we leverage location-based and credit card transaction data in spatial analyses to proactively evaluate the recovery of communities and the impacts of floods on businesses. These analyses show that the significance of core capabilities of the smart flood resilience framework in helping emergency managers, city planners, public officials, responders, and volunteers to better cope with the impacts of catastrophic flooding events.

physics.soc-ph↗

arXiv2021

Layer-controlled Ferromagnetism in Atomically Thin CrSiTe$_3$ Flakes

The research on two-dimensional (2D) van der Waals (vdW) ferromagnets has promoted the development of ultrahigh-density and nanoscale data storage. However, intrinsic ferromagnetism in layered magnets is always subject to many factors, such as stacking orders, interlayer couplings, and the number of layers. Here, we report a magnetic transition from soft to hard ferromagnetic behaviors as the thickness of CrSiTe$_3$ flakes decreases down to several nanometers. Phenomenally, in contrast to the negligible hysteresis loop in the bulk counterparts, atomically thin CrSiTe$_3$ shows a rectangular loop with finite magnetization and coercivity as thickness decreases down to ~8 nm, indicative of a single-domain and out-of-plane ferromagnetic order. We find that the stray field is weakened with decreasing thickness, which suppresses the formation of the domain wall. In addition, thickness-dependent ferromagnetic properties also reveal a crossover from 3 dimensional to 2 dimensional Ising ferromagnets at a ~7 nm thickness of CrSiTe$_3$, accompanied by a drop of the Curie temperature from 33 K for bulk to ~17 K for 4 nm sample.

cond-mat.mtrl-sci↗

arXiv2021

Identifiable Generative Models for Missing Not at Random Data Imputation

Real-world datasets often have missing values associated with complex generative processes, where the cause of the missingness may not be fully observed. This is known as missing not at random (MNAR) data. However, many imputation methods do not take into account the missingness mechanism, resulting in biased imputation values when MNAR data is present. Although there are a few methods that have considered the MNAR scenario, their model's identifiability under MNAR is generally not guaranteed. That is, model parameters can not be uniquely determined even with infinite data samples, hence the imputation results given by such models can still be biased. This issue is especially overlooked by many modern deep generative models. In this work, we fill in this gap by systematically analyzing the identifiability of generative models under MNAR. Furthermore, we propose a practical deep generative model which can provide identifiability guarantees under mild assumptions, for a wide range of MNAR mechanisms. Our method demonstrates a clear advantage for tasks on both synthetic data and multiple real-world scenarios with MNAR data.

cs.LG↗