arXiv subjects

Li Ma

Publications and source records attributed to Li Ma.

At least 109 records · Page 6Linked to original sources

arXiv2023

Alteration Detection of Tensor Dependence Structure via Sparsity-Exploited Reranking Algorithm

Tensor-valued data arise frequently from a wide variety of scientific applications, and many among them can be translated into an alteration detection problem of tensor dependence structures. In this article, we formulate the problem under the popularly adopted tensor-normal distributions and aim at two-sample correlation/partial correlation comparisons of tensor-valued observations. Through decorrelation and centralization, a separable covariance structure is employed to pool sample information from different tensor modes to enhance the power of the test. Additionally, we propose a novel Sparsity-Exploited Reranking Algorithm (SERA) to further improve the multiple testing efficiency. The algorithm is approached through reranking of the p-values derived from the primary test statistics, by incorporating a carefully constructed auxiliary tensor sequence. Besides the tensor framework, SERA is also generally applicable to a wide range of two-sample large-scale inference problems with sparsity structures, and is of independent interest. The asymptotic properties of the proposed test are derived and the algorithm is shown to control the false discovery at the pre-specified level. We demonstrate the efficacy of the proposed method through intensive simulations and two scientific applications.

stat.ME↗

arXiv2023

A Decorrelating and Debiasing Approach to Simultaneous Inference for High-Dimensional Confounded Models

Motivated by the simultaneous association analysis with the presence of latent confounders, this paper studies the large-scale hypothesis testing problem for the high-dimensional confounded linear models with both non-asymptotic and asymptotic false discovery control. Such model covers a wide range of practical settings where both the response and the predictors may be confounded. In the presence of the high-dimensional predictors and the unobservable confounders, the simultaneous inference with provable guarantees becomes highly challenging, and the unknown strong dependence among the confounded covariates makes the challenge even more pronounced. This paper first introduces a decorrelating procedure that shrinks the confounding effect and weakens the correlations among the predictors, then performs debiasing under the decorrelated design based on some biased initial estimator. Following that, an asymptotic normality result for the debiased estimator is established and standardized test statistics are then constructed. Furthermore, a simultaneous inference procedure is proposed to identify significant associations, and both the finite-sample and asymptotic false discovery bounds are provided. The non-asymptotic result is general and model-free, and is of independent interest. We also prove that, under minimal signal strength condition, all associations can be successfully detected with probability tending to one. Simulation and real data studies are carried out to evaluate the performance of the proposed approach and compare it with other competing methods.

stat.ME↗

arXiv2023

NAPA: Neighborhood-Assisted and Posterior-Adjusted Two-sample Inference

Two-sample multiple testing problems of sparse spatial data are frequently arising in a variety of scientific applications. In this article, we develop a novel neighborhood-assisted and posterior-adjusted (NAPA) approach to incorporate both the spatial smoothness and sparsity type side information to improve the power of the test while controlling the false discovery of multiple testing. We translate the side information into a set of weights to adjust the $p$-values, where the spatial pattern is encoded by the ordering of the locations, and the sparsity structure is encoded by a set of auxiliary covariates. We establish the theoretical properties of the proposed test, including the guaranteed power improvement over some state-of-the-art alternative tests, and the asymptotic false discovery control. We demonstrate the efficacy of the test through intensive simulations and two neuroimaging applications.

stat.ME↗

arXiv2023

Unsupervised tree boosting for learning probability distributions

We propose an unsupervised tree boosting algorithm for inferring the underlying sampling distribution of an i.i.d. sample based on fitting additive tree ensembles in a fashion analogous to supervised tree boosting. Integral to the algorithm is a new notion of "addition" on probability distributions that leads to a coherent notion of "residualization", i.e., subtracting a probability distribution from an observation to remove the distributional structure from the sampling distribution of the latter. We show that these notions arise naturally for univariate distributions through cumulative distribution function (CDF) transforms and compositions due to several "group-like" properties of univariate CDFs. While the traditional multivariate CDF does not preserve these properties, a new definition of multivariate CDF can restore these properties, thereby allowing the notions of "addition" and "residualization" to be formulated for multivariate settings as well. This then gives rise to the unsupervised boosting algorithm based on forward-stagewise fitting of an additive tree ensemble, which sequentially reduces the Kullback-Leibler divergence from the truth. The algorithm allows analytic evaluation of the fitted density and outputs a generative model that can be readily sampled from. We enhance the algorithm with scale-dependent shrinkage and a two-stage strategy that separately fits the marginals and the copula. The algorithm then performs competitively to state-of-the-art deep-learning approaches in multivariate density estimation on multiple benchmark data sets.

stat.ME↗

arXiv2023

OpenAPMax: Abnormal Patterns-based Model for Real-World Alzheimer's Disease Diagnosis

Alzheimer's disease (AD) cannot be reversed, but early diagnosis will significantly benefit patients' medical treatment and care. In recent works, AD diagnosis has the primary assumption that all categories are known a prior -- a closed-set classification problem, which contrasts with the open-set recognition problem. This assumption hinders the application of the model in natural clinical settings. Although many open-set recognition technologies have been proposed in other fields, they are challenging to use for AD diagnosis directly since 1) AD is a degenerative disease of the nervous system with similar symptoms at each stage, and it is difficult to distinguish from its pre-state, and 2) diversified strategies for AD diagnosis are challenging to model uniformly. In this work, inspired by the concerns of clinicians during diagnosis, we propose an open-set recognition model, OpenAPMax, based on the anomaly pattern to address AD diagnosis in real-world settings. OpenAPMax first obtains the abnormal pattern of each patient relative to each known category through statistics or a literature search, clusters the patients' abnormal pattern, and finally, uses extreme value theory (EVT) to model the distance between each patient's abnormal pattern and the center of their category and modify the classification probability. We evaluate the performance of the proposed method with recent open-set recognition, where we obtain state-of-the-art results.

cs.LG↗

arXiv2023

OpenClinicalAI: An Open and Dynamic Model for Alzheimer's Disease Diagnosis

Although Alzheimer's disease (AD) cannot be reversed or cured, timely diagnosis can significantly reduce the burden of treatment and care. Current research on AD diagnosis models usually regards the diagnosis task as a typical classification task with two primary assumptions: 1) All target categories are known a priori; 2) The diagnostic strategy for each patient is consistent, that is, the number and type of model input data for each patient are the same. However, real-world clinical settings are open, with complexity and uncertainty in terms of both subjects and the resources of the medical institutions. This means that diagnostic models may encounter unseen disease categories and need to dynamically develop diagnostic strategies based on the subject's specific circumstances and available medical resources. Thus, the AD diagnosis task is tangled and coupled with the diagnosis strategy formulation. To promote the application of diagnostic systems in real-world clinical settings, we propose OpenClinicalAI for direct AD diagnosis in complex and uncertain clinical settings. This is the first powerful end-to-end model to dynamically formulate diagnostic strategies and provide diagnostic results based on the subject's conditions and available medical resources. OpenClinicalAI combines reciprocally coupled deep multiaction reinforcement learning (DMARL) for diagnostic strategy formulation and multicenter meta-learning (MCML) for open-set recognition. The experimental results show that OpenClinicalAI achieves better performance and fewer clinical examinations than the state-of-the-art model. Our method provides an opportunity to embed the AD diagnostic system into the current health care system to cooperate with clinicians to improve current health care.

cs.LG↗

arXiv2023

Picking Up Quantization Steps for Compressed Image Classification

The sensitivity of deep neural networks to compressed images hinders their usage in many real applications, which means classification networks may fail just after taking a screenshot and saving it as a compressed file. In this paper, we argue that neglected disposable coding parameters stored in compressed files could be picked up to reduce the sensitivity of deep neural networks to compressed images. Specifically, we resort to using one of the representative parameters, quantization steps, to facilitate image classification. Firstly, based on quantization steps, we propose a novel quantization aware confidence (QAC), which is utilized as sample weights to reduce the influence of quantization on network training. Secondly, we utilize quantization steps to alleviate the variance of feature distributions, where a quantization aware batch normalization (QABN) is proposed to replace batch normalization of classification networks. Extensive experiments show that the proposed method significantly improves the performance of classification networks on CIFAR-10, CIFAR-100, and ImageNet. The code is released on https://github.com/LiMaPKU/QSAM.git

cs.CV↗

arXiv2023

The global solution of the minimal surface flow and translating surfaces

In this paper, we study evolved surfaces over convex planar domains which are evolving by the minimal surface flow $$u_{t}= div\left(\frac{Du}{\sqrt{1+|Du|^2}}\right)-H(x,Du).$$ Here, we specify the angle of contact of the evolved surface to the boundary cylinder. The interesting question is to find translating solitons of the form $u(x,t)=ωt+w(x)$ where $ω\in \mathbb R$. Under an angle condition, we can prove the a priori estimate holds true for the translating solitons (i.e., translator), which makes the solitons exist. We can prove for suitable condition on $H(x,p)$ that there is the global solution of the minimal surface flow. Then we show, provided the soliton exists, that the global solutions converge to some translator.

math.DG↗

arXiv2023

3D Video Loops from Asynchronous Input

Looping videos are short video clips that can be looped endlessly without visible seams or artifacts. They provide a very attractive way to capture the dynamism of natural scenes. Existing methods have been mostly limited to 2D representations. In this paper, we take a step forward and propose a practical solution that enables an immersive experience on dynamic 3D looping scenes. The key challenge is to consider the per-view looping conditions from asynchronous input while maintaining view consistency for the 3D representation. We propose a novel sparse 3D video representation, namely Multi-Tile Video (MTV), which not only provides a view-consistent prior, but also greatly reduces memory usage, making the optimization of a 4D volume tractable. Then, we introduce a two-stage pipeline to construct the 3D looping MTV from completely asynchronous multi-view videos with no time overlap. A novel looping loss based on video temporal retargeting algorithms is adopted during the optimization to loop the 3D scene. Experiments of our framework have shown promise in successfully generating and rendering photorealistic 3D looping videos in real time even on mobile devices. The code, dataset, and live demos are available in https://limacv.github.io/VideoLoop3D_web/.

cs.CV↗

arXiv2023

Practical Continuous-variable Quantum Key Distribution with Feasible Optimization Parameters

Continuous-variable quantum key distribution (CV-QKD) offers an approach to achieve a potential high secret key rate (SKR) in metropolitan areas. There are several challenges in developing a practical CV-QKD system from the laboratory to the real world. One of the most significant points is that it is really hard to adapt different practical optical fiber conditions for CV-QKD systems with unified hardware. Thus, how to improve the performance of practical CV-QKD systems in the field without modification of the hardware is very important. Here, a systematic optimization method, combining the modulation variance and error correction matrix optimization, is proposed to improve the performance of a practical CV-QKD system with a restricted capacity of postprocessing. The effect of restricted postprocessing capacity on the SKR is modeled as a nonlinear programming problem with modulation variance as an optimization parameter, and the selection of an optimal error correction matrix is studied under the same scheme. The results show that the SKR of a CV-QKD system can be improved by 24% and 200% compared with previous frequently used optimization methods theoretically with a transmission distance of 50 km. Furthermore, the experimental results verify the feasibility and robustness of the proposed method, where the achieved optimal SKR achieved practically deviates <1.6% from the theoretical optimal value. Our results pave the way to deploy high-performance CV-QKD in the real world.

quant-ph↗

arXiv2022

Performance analysis for OFDM-based multi-carrier continuous-variable quantum key distribution with arbitrary modulation protocol

Multi-carrier continuous-variable quantum key distribution (CV-QKD) is considered to be a promising way to boost the secret key rate (SKR) over the existing single-carrier CV-QKD scheme. However, the extra excess noise induced in the imperfect multi-carrier quantum state preparation process of N subcarriers will limit the performance of the system. Here, a systematic modulation noise model is proposed for the multi-carrier CV-QKD based on the orthogonal frequency division multiplexing (OFDM). Subsequently, the performance of multi-carrier CV-QKD with arbitrary modulation protocol (e.g. QPSK, 256QAM and Gaussian modulation protocol) can be quantitatively evaluated by combining the security analysis method of the single-carrier CV-QKD. Under practical system parameters, the simulation results show that the SKR of the multi-carrier CV-QKD can still be significantly improved by increasing the carrier number N even with imperfect practical modulations. Specifically, the total SKR of multi-carrier CV-QKD can be optimized by carefully choosing N. The proposed model provides a feasible theoretical framework for the future multi-carrier CV-QKD experimental implementation.

quant-ph↗

arXiv2022

Neural Parameterization for Dynamic Human Head Editing

Implicit radiance functions emerged as a powerful scene representation for reconstructing and rendering photo-realistic views of a 3D scene. These representations, however, suffer from poor editability. On the other hand, explicit representations such as polygonal meshes allow easy editing but are not as suitable for reconstructing accurate details in dynamic human heads, such as fine facial features, hair, teeth, and eyes. In this work, we present Neural Parameterization (NeP), a hybrid representation that provides the advantages of both implicit and explicit methods. NeP is capable of photo-realistic rendering while allowing fine-grained editing of the scene geometry and appearance. We first disentangle the geometry and appearance by parameterizing the 3D geometry into 2D texture space. We enable geometric editability by introducing an explicit linear deformation blending layer. The deformation is controlled by a set of sparse key points, which can be explicitly and intuitively displaced to edit the geometry. For appearance, we develop a hybrid 2D texture consisting of an explicit texture map for easy editing and implicit view and time-dependent residuals to model temporal and view variations. We compare our method to several reconstruction and editing baselines. The results show that the NeP achieves almost the same level of rendering accuracy while maintaining high editability.

cs.CV↗

arXiv2022

Microwave-optical double resonance in a erbium-doped whispering-gallery-mode resonator

We showcase an erbium-doped whispering-gallery-mode resonator with optical modes that display intrinsic quality factors better than $10^8$ (linewidths less than 2 MHz), and coupling strengths to collective erbium transitions of up to 2$π\times$1.2 GHz - enough to reach the ensemble strong coupling regime. Our optical cavity sits inside a microwave resonator, allowing us to probe the spin transition which is tuned by an external magnetic field. We show a modified optically detected magnetic resonance measurement that measures population transfer by a change in coupling strength rather than absorption coefficient. This modification was enabled by the strong coupling to our modes, and allows us to optically probe the spin transition detuned by more than the inhomogeneous linewidth. We contrast this measurement with electron paramagnetic resonance to experimentally show that our optical modes are confined in a region of large microwave magnetic field and we explore how such a geometry could be used for coherent microwave-optical transduction.

quant-ph↗

arXiv2022

Fine-Grained Modeling and Optimization for Intelligent Resource Management in Big Data Processing

Big data processing at the production scale presents a highly complex environment for resource optimization (RO), a problem crucial for meeting performance goals and budgetary constraints of analytical users. The RO problem is challenging because it involves a set of decisions (the partition count, placement of parallel instances on machines, and resource allocation to each instance), requires multi-objective optimization (MOO), and is compounded by the scale and complexity of big data systems while having to meet stringent time constraints for scheduling. This paper presents a MaxCompute-based integrated system to support multi-objective resource optimization via fine-grained instance-level modeling and optimization. We propose a new architecture that breaks RO into a series of simpler problems, new fine-grained predictive models, and novel optimization methods that exploit these models to make effective instance-level recommendations in a hierarchical MOO framework. Evaluation using production workloads shows that our new RO system could reduce 37-72% latency and 43-78% cost at the same time, compared to the current optimizer and scheduler, while running in 0.02-0.23s.

cs.DB↗

arXiv2022

High-throughput decoder of quasi-cyclic LDPC codes with limited precision for continuous-variable quantum key distribution systems

More than Mbps secret key rate was demonstrated for continuous-variable quantum key distribution (CV-QKD) systems, but real-time postprocessing is not allowed, which is restricted by the throughput of the error correction decoding in postprocessing. In this paper, a high-throughput FPGA-based quasi-cyclic LDPC decoder is proposed and implemented to support Mbps real-time secret key rate generation for CV-QKD for the first time. A residual bit error correction algorithm is used to solve the problem of high frame errors rate (FER) caused by the limited precision of the decoder. Specifically, real-time high-speed decoding for CV-QKD systems with typical code rates 0.2 and 0.1 is implemented on a commercial FPGA, and two throughputs of 360.92Mbps and 194.65Mbps are achieved, respectively, which can support 17.97 Mbps and 2.48 Mbps real-time generation of secret key rates under typical transmission distances of 25km and 50km, correspondingly. The proposed method paves the way for high-rate real-time CV-QKD deployment in secure metropolitan area network.

quant-ph↗

arXiv2022

Microbiome subcommunity learning with logistic-tree normal latent Dirichlet allocation

Mixed-membership (MM) models such as Latent Dirichlet Allocation (LDA) have been applied to microbiome compositional data to identify latent subcommunities of microbial species. These subcommunities are informative for understanding the biological interplay of microbes and for predicting health outcomes. However, microbiome compositions typically display substantial cross-sample heterogeneities in subcommunity compositions -- that is, the variability in the proportions of microbes in shared subcommunities across samples -- which is not accounted for in prior analyses. As a result, LDA can produce inference which is highly sensitive to the specification of the number of subcommunities and often divides a single subcommunity into multiple artificial ones. To address this limitation, we incorporate the logistic-tree normal (LTN) model into LDA to form a new MM model. This model allows cross-sample variation in the composition of each subcommunity around some "centroid" composition that defines the subcommunity. Incorporation of auxiliary Pólya-Gamma variables enables a computationally efficient collapsed blocked Gibbs sampler to carry out Bayesian inference under this model. By accounting for such heterogeneity, our new model restores the robustness of the inference in the specification of the number of subcommunities and allows meaningful subcommunities to be identified.

stat.AP↗

arXiv2022

Deblur-NeRF: Neural Radiance Fields from Blurry Images

Neural Radiance Field (NeRF) has gained considerable attention recently for 3D scene reconstruction and novel view synthesis due to its remarkable synthesis quality. However, image blurriness caused by defocus or motion, which often occurs when capturing scenes in the wild, significantly degrades its reconstruction quality. To address this problem, We propose Deblur-NeRF, the first method that can recover a sharp NeRF from blurry input. We adopt an analysis-by-synthesis approach that reconstructs blurry views by simulating the blurring process, thus making NeRF robust to blurry inputs. The core of this simulation is a novel Deformable Sparse Kernel (DSK) module that models spatially-varying blur kernels by deforming a canonical sparse kernel at each spatial location. The ray origin of each kernel point is jointly optimized, inspired by the physical blurring process. This module is parameterized as an MLP that has the ability to be generalized to various blur types. Jointly optimizing the NeRF and the DSK module allows us to restore a sharp NeRF. We demonstrate that our method can be used on both camera motion blur and defocus blur: the two most common types of blur in real scenes. Evaluation results on both synthetic and real-world data show that our method outperforms several baselines. The synthetic and real datasets along with the source code is publicly available at https://limacv.github.io/deblurnerf/

cs.CV↗

arXiv2022

TVConv: Efficient Translation Variant Convolution for Layout-aware Visual Processing

As convolution has empowered many smart applications, dynamic convolution further equips it with the ability to adapt to diverse inputs. However, the static and dynamic convolutions are either layout-agnostic or computation-heavy, making it inappropriate for layout-specific applications, e.g., face recognition and medical image segmentation. We observe that these applications naturally exhibit the characteristics of large intra-image (spatial) variance and small cross-image variance. This observation motivates our efficient translation variant convolution (TVConv) for layout-aware visual processing. Technically, TVConv is composed of affinity maps and a weight-generating block. While affinity maps depict pixel-paired relationships gracefully, the weight-generating block can be explicitly overparameterized for better training while maintaining efficient inference. Although conceptually simple, TVConv significantly improves the efficiency of the convolution and can be readily plugged into various network architectures. Extensive experiments on face recognition show that TVConv reduces the computational cost by up to 3.1x and improves the corresponding throughput by 2.3x while maintaining a high accuracy compared to the depthwise convolution. Moreover, for the same computation cost, we boost the mean accuracy by up to 4.21%. We also conduct experiments on the optic disc/cup segmentation task and obtain better generalization performance, which helps mitigate the critical data scarcity issue. Code is available at https://github.com/JierunChen/TVConv.

cs.CV↗