arXiv subjects

Yan Wang

Publications and source records attributed to Yan Wang.

At least 721 records · Page 40Linked to original sources

arXiv2022

Fast Calibration for Computer Models with Massive Physical Observations

Computer model calibration is a crucial step in building a reliable computer model. In the face of massive physical observations, a fast estimation for the calibration parameters is urgently needed. To alleviate the computational burden, we design a two-step algorithm to estimate the calibration parameters by employing the subsampling techniques. Compared with the current state-of-the-art calibration methods, the complexity of the proposed algorithm is greatly reduced without sacrificing too much accuracy. We prove the consistency and asymptotic normality of the proposed estimator. The form of the variance of the proposed estimation is also presented, which provides a natural way to quantify the uncertainty of the calibration parameters. The obtained results of two numerical simulations and two real-case studies demonstrate the advantages of the proposed method.

math.ST↗

arXiv2022

FedVMR: A New Federated Learning method for Video Moment Retrieval

Despite the great success achieved, existing video moment retrieval (VMR) methods are developed under the assumption that data are centralizedly stored. However, in real-world applications, due to the inherent nature of data generation and privacy concerns, data are often distributed on different silos, bringing huge challenges to effective large-scale training. In this work, we try to overcome above limitation by leveraging the recent success of federated learning. As the first that is explored in VMR field, the new task is defined as video moment retrieval with distributed data. Then, a novel federated learning method named FedVMR is proposed to facilitate large-scale and secure training of VMR models in decentralized environment. Experiments on benchmark datasets demonstrate its effectiveness. This work is the very first attempt to enable safe and efficient VMR training in decentralized scene, which is hoped to pave the way for further study in the related research field.

cs.CV↗

arXiv2022

Neural Surface Reconstruction of Dynamic Scenes with Monocular RGB-D Camera

We propose Neural-DynamicReconstruction (NDR), a template-free method to recover high-fidelity geometry and motions of a dynamic scene from a monocular RGB-D camera. In NDR, we adopt the neural implicit function for surface representation and rendering such that the captured color and depth can be fully utilized to jointly optimize the surface and deformations. To represent and constrain the non-rigid deformations, we propose a novel neural invertible deforming network such that the cycle consistency between arbitrary two frames is automatically satisfied. Considering that the surface topology of dynamic scene might change over time, we employ a topology-aware strategy to construct the topology-variant correspondence for the fused frames. NDR also further refines the camera poses in a global optimization manner. Experiments on public datasets and our collected dataset demonstrate that NDR outperforms existing monocular dynamic reconstruction methods.

cs.CV↗

arXiv2022

Correcting the Sub-optimal Bit Allocation

In this paper, we investigate the problem of bit allocation in Neural Video Compression (NVC). First, we reveal that a recent bit allocation approach claimed to be optimal is, in fact, sub-optimal due to its implementation. Specifically, we find that its sub-optimality lies in the improper application of semi-amortized variational inference (SAVI) on latent with non-factorized variational posterior. Then, we show that the corrected version of SAVI on non-factorized latent requires recursively applying back-propagating through gradient ascent, based on which we derive the corrected optimal bit allocation algorithm. Due to the computational in-feasibility of the corrected bit allocation, we design an efficient approximation to make it practical. Empirical results show that our proposed correction significantly improves the incorrect bit allocation in terms of R-D performance and bitrate error, and outperforms all other bit allocation methods by a large margin. The source code is provided in the supplementary material.

cs.CV↗

arXiv2022

Spatial Moment Pooling Improves Neural Image Assessment

In recent years, there has been widespread attention drawn to convolutional neural network (CNN) based blind image quality assessment (IQA). A large number of works start by extracting deep features from CNN. Then, those features are processed through spatial average pooling (SAP) and fully connected layers to predict quality. Inspired by full reference IQA and texture features, in this paper, we extend SAP ($1^{st}$ moment) into spatial moment pooling (SMP) by incorporating higher order moments (such as variance, skewness). Moreover, we provide learning friendly normalization to circumvent numerical issue when computing gradients of higher moments. Experimental results suggest that simply upgrading SAP to SMP significantly enhances CNN-based blind IQA methods and achieves state of the art performance.

cs.CV↗

arXiv2022

Multi-Sample Training for Neural Image Compression

This paper considers the problem of lossy neural image compression (NIC). Current state-of-the-art (sota) methods adopt uniform posterior to approximate quantization noise, and single-sample pathwise estimator to approximate the gradient of evidence lower bound (ELBO). In this paper, we propose to train NIC with multiple-sample importance weighted autoencoder (IWAE) target, which is tighter than ELBO and converges to log likelihood as sample size increases. First, we identify that the uniform posterior of NIC has special properties, which affect the variance and bias of pathwise and score function estimators of the IWAE target. Moreover, we provide insights on a commonly adopted trick in NIC from gradient variance perspective. Based on those analysis, we further propose multiple-sample NIC (MS-NIC), an enhanced IWAE target for NIC. Experimental results demonstrate that it improves sota NIC methods. Our MS-NIC is plug-and-play, and can be easily extended to other neural compression tasks.

cs.CV↗

arXiv2022

Black-Box Dissector: Towards Erasing-based Hard-Label Model Stealing Attack

Previous studies have verified that the functionality of black-box models can be stolen with full probability outputs. However, under the more practical hard-label setting, we observe that existing methods suffer from catastrophic performance degradation. We argue this is due to the lack of rich information in the probability prediction and the overfitting caused by hard labels. To this end, we propose a novel hard-label model stealing method termed \emph{black-box dissector}, which consists of two erasing-based modules. One is a CAM-driven erasing strategy that is designed to increase the information capacity hidden in hard labels from the victim model. The other is a random-erasing-based self-knowledge distillation module that utilizes soft labels from the substitute model to mitigate overfitting. Extensive experiments on four widely-used datasets consistently demonstrate that our method outperforms state-of-the-art methods, with an improvement of at most $8.27\%$. We also validate the effectiveness and practical potential of our method on real-world APIs and defense methods. Furthermore, our method promotes other downstream tasks, \emph{i.e.}, transfer adversarial attacks.

cs.CV↗

arXiv2022

A Contrastive Framework for Neural Text Generation

Text generation is of great importance to many natural language processing applications. However, maximization-based decoding methods (e.g. beam search) of neural language models often lead to degenerate solutions -- the generated text is unnatural and contains undesirable repetitions. Existing approaches introduce stochasticity via sampling or modify training objectives to decrease probabilities of certain tokens (e.g., unlikelihood training). However, they often lead to solutions that lack coherence. In this work, we show that an underlying reason for model degeneration is the anisotropic distribution of token representations. We present a contrastive solution: (i) SimCTG, a contrastive training objective to calibrate the model's representation space, and (ii) a decoding method -- contrastive search -- to encourage diversity while maintaining coherence in the generated text. Extensive experiments and analyses on three benchmarks from two languages demonstrate that our proposed approach significantly outperforms current state-of-the-art text generation methods as evaluated by both human and automatic metrics.

cs.CL↗

arXiv2022

Investigations on the charmless decays of $X(3872)$ in intermediate meson loops model

The charmless decay processes of $X(3872)$ provide us a good platform to study the nature and the decay mechanism of $X(3872)$. Based on a molecular nature of $X(3872)$ as a $\bar{D}D^*$ bound state, we have investigated the charmless decays $X(3872) \to VV$ and $VP$ via intermediate $D^*{\bar D} +c.c.$ meson loops, where $V$ and $P$ stand for light vector and pseudoscalar mesons, respectively. We discuss three cases, i.e., pure neutral components ($θ=0$), isospin singlet ($θ=π/4$) and neutral components dominant ($θ= π/6$), where $θ$ is a phase angle describing the proportion of neutral and charged constituents. The proportion of neutral and charged constituent have an influence on the decay widths of $X(3872) \to VV$ and $VP$. With the coupling constant of $X(3872)$ to the $\bar{D}D^*$ channel obtained under the molecule ansatz of $X(3872)$ resonance, the predicted decay widths of $X(3872)\rightarrow VV$ are about tens of keVs, while the decay width can reach a few hundreds of keVs for $X(3872)\to VP$. The dependence of these ratios between different decay modes of $X(3872)\to VV$ and $X(3872)\to VP$ to the mixing angle $θ$ is also investigated. It is expected that the theoretical calculations here can be tested by future experiments.

hep-ph↗

arXiv2022

HiFuse: Hierarchical Multi-Scale Feature Fusion Network for Medical Image Classification

Medical image classification has developed rapidly under the impetus of the convolutional neural network (CNN). Due to the fixed size of the receptive field of the convolution kernel, it is difficult to capture the global features of medical images. Although the self-attention-based Transformer can model long-range dependencies, it has high computational complexity and lacks local inductive bias. Much research has demonstrated that global and local features are crucial for image classification. However, medical images have a lot of noisy, scattered features, intra-class variation, and inter-class similarities. This paper proposes a three-branch hierarchical multi-scale feature fusion network structure termed as HiFuse for medical image classification as a new method. It can fuse the advantages of Transformer and CNN from multi-scale hierarchies without destroying the respective modeling so as to improve the classification accuracy of various medical images. A parallel hierarchy of local and global feature blocks is designed to efficiently extract local features and global representations at various semantic scales, with the flexibility to model at different scales and linear computational complexity relevant to image size. Moreover, an adaptive hierarchical feature fusion block (HFF block) is designed to utilize the features obtained at different hierarchical levels comprehensively. The HFF block contains spatial attention, channel attention, residual inverted MLP, and shortcut to adaptively fuse semantic information between various scale features of each branch. The accuracy of our proposed model on the ISIC2018 dataset is 7.6% higher than baseline, 21.5% on the Covid-19 dataset, and 10.4% on the Kvasir dataset. Compared with other advanced models, the HiFuse model performs the best. Our code is open-source and available from https://github.com/huoxiangzuo/HiFuse.

eess.IV↗

arXiv2022

S$^3$R: Self-supervised Spectral Regression for Hyperspectral Histopathology Image Classification

Benefited from the rich and detailed spectral information in hyperspectral images (HSI), HSI offers great potential for a wide variety of medical applications such as computational pathology. But, the lack of adequate annotated data and the high spatiospectral dimensions of HSIs usually make classification networks prone to overfit. Thus, learning a general representation which can be transferred to the downstream tasks is imperative. To our knowledge, no appropriate self-supervised pre-training method has been designed for histopathology HSIs. In this paper, we introduce an efficient and effective Self-supervised Spectral Regression (S$^3$R) method, which exploits the low rank characteristic in the spectral domain of HSI. More concretely, we propose to learn a set of linear coefficients that can be used to represent one band by the remaining bands via masking out these bands. Then, the band is restored by using the learned coefficients to reweight the remaining bands. Two pre-text tasks are designed: (1)S$^3$R-CR, which regresses the linear coefficients, so that the pre-trained model understands the inherent structures of HSIs and the pathological characteristics of different morphologies; (2)S$^3$R-BR, which regresses the missing band, making the model to learn the holistic semantics of HSIs. Compared to prior arts i.e., contrastive learning methods, which focuses on natural images, S$^3$R converges at least 3 times faster, and achieves significant improvements up to 14% in accuracy when transferring to HSI classification tasks.

cs.CV↗

arXiv2022

Flexible Neural Image Compression via Code Editing

Neural image compression (NIC) has outperformed traditional image codecs in rate-distortion (R-D) performance. However, it usually requires a dedicated encoder-decoder pair for each point on R-D curve, which greatly hinders its practical deployment. While some recent works have enabled bitrate control via conditional coding, they impose strong prior during training and provide limited flexibility. In this paper we propose Code Editing, a highly flexible coding method for NIC based on semi-amortized inference and adaptive quantization. Our work is a new paradigm for variable bitrate NIC. Furthermore, experimental results show that our method surpasses existing variable-rate methods, and achieves ROI coding and multi-distortion trade-off with a single decoder.

eess.IV↗

arXiv2022

Remote entanglement distribution in a quantum network via multinode indistinguishability of photons

Quantum networking relies on entanglement distribution between distant nodes, typically realized by swapping procedures. However, entanglement swapping is a demanding task in practice, mainly because of limited effectiveness of entangled photon sources and Bell-state measurements necessary to realize the process. Here we experimentally activate a remote distribution of two-photon polarization entanglement superseding the need for initial entangled pairs and traditional Bell-state measurements. This alternative procedure is accomplished thanks to the controlled spatial indistinguishability of four independent photons in three separated nodes of the network, which enables us to perform localized product-state measurements in the central node acting as a trigger. This experiment proves that the inherent indistinguishability of identical particles supplies new standards for feasible quantum communication in multinode photonic quantum networks.

quant-ph↗

arXiv2022

Approximation of Images via Generalized Higher Order Singular Value Decomposition over Finite-dimensional Commutative Semisimple Algebra

Low-rank approximation of images via singular value decomposition is well-received in the era of big data. However, singular value decomposition (SVD) is only for order-two data, i.e., matrices. It is necessary to flatten a higher order input into a matrix or break it into a series of order-two slices to tackle higher order data such as multispectral images and videos with the SVD. Higher order singular value decomposition (HOSVD) extends the SVD and can approximate higher order data using sums of a few rank-one components. We consider the problem of generalizing HOSVD over a finite dimensional commutative algebra. This algebra, referred to as a t-algebra, generalizes the field of complex numbers. The elements of the algebra, called t-scalars, are fix-sized arrays of complex numbers. One can generalize matrices and tensors over t-scalars and then extend many canonical matrix and tensor algorithms, including HOSVD, to obtain higher-performance versions. The generalization of HOSVD is called THOSVD. Its performance of approximating multi-way data can be further improved by an alternating algorithm. THOSVD also unifies a wide range of principal component analysis algorithms. To exploit the potential of generalized algorithms using t-scalars for approximating images, we use a pixel neighborhood strategy to convert each pixel to "deeper-order" t-scalar. Experiments on publicly available images show that the generalized algorithm over t-scalars, namely THOSVD, compares favorably with its canonical counterparts.

cs.LG↗

arXiv2022

RIBAC: Towards Robust and Imperceptible Backdoor Attack against Compact DNN

Recently backdoor attack has become an emerging threat to the security of deep neural network (DNN) models. To date, most of the existing studies focus on backdoor attack against the uncompressed model; while the vulnerability of compressed DNNs, which are widely used in the practical applications, is little exploited yet. In this paper, we propose to study and develop Robust and Imperceptible Backdoor Attack against Compact DNN models (RIBAC). By performing systematic analysis and exploration on the important design knobs, we propose a framework that can learn the proper trigger patterns, model parameters and pruning masks in an efficient way. Thereby achieving high trigger stealthiness, high attack success rate and high model efficiency simultaneously. Extensive evaluations across different datasets, including the test against the state-of-the-art defense mechanisms, demonstrate the high robustness, stealthiness and model efficiency of RIBAC. Code is available at https://github.com/huyvnphan/ECCV2022-RIBAC

cs.CR↗

arXiv2022

Nested Bayesian Optimization for Computer Experiments

Computer experiments can emulate the physical systems, help computational investigations, and yield analytic solutions. They have been widely employed with many engineering applications (e.g., aerospace, automotive, energy systems. Conventional Bayesian optimization did not incorporate the nested structures in computer experiments. This paper proposes a novel nested Bayesian optimization for complex computer experiments with multi-step or hierarchical characteristics. We prove the theoretical properties of nested outputs given two cases: Gaussian or non-Gaussian. The closed forms of nested expected improvement are derived. We also propose the computational algorithms for nested Bayesian optimization. Three numerical studies show that the proposed nested Bayesian optimization outperforms the five benchmark Bayesian optimization methods ignoring the intermediate outputs of the inner computer code. The case study shows that the nested Bayesian optimization can efficiently minimize the residual stress during composite structures assembly and avoid convergence to the local optimum.

stat.ME↗

arXiv2022

The saturated pairwise interaction Gibbs point process as a joint species distribution model

In an effort to effectively model observed patterns in the spatial configuration of individuals of multiple species in nature, we introduce the saturated pairwise interaction Gibbs point process. Its main strength lies in its ability to model both attraction and repulsion within and between species, over different scales. As such, it is particularly well-suited to the study of associations in complex ecosystems. Based on the existing literature, we provide an easy to implement fitting procedure as well as a technique to make inference for the model parameters. We also prove that under certain hypotheses the point process is locally stable, which allows us to use the well-known `coupling from the past' algorithm to draw samples from the model. Different numerical experiments show the robustness of the model. We study three different ecological datasets, demonstrating in each one that our model helps disentangle competing ecological effects on species' distribution.

stat.ME↗

arXiv2022

On constraining the mesoscale eddy energy dissipation time-scale

A physically plausible lower bound on the spatially varying geostrophic mesoscale eddy energy dissipation time-scale within the ocean, related to the geographical energy transfer rate out of the geostrophic mesoscales, is provided by means of a simple and computational inexpensive inverse calculation. Data diagnosed from a high resolution global configuration ocean simulation is supplied to a parameterized model of the geostrophic mesoscale eddy energy, from which the dissipation time-scale results as a solution to an optimization calculation. We find that the dissipation time-scale is shortest in the Southern Ocean, in the Western Boundary Currents, and on the western boundaries, consistent with the expectation that these regions are notable sites of baroclinic activity with processes leading to energy transfer out of the geostrophic mesoscales. Although our solution should be interpreted as a lower bound given the assumptions going into the calculation, it serves as an important physically consistent base line reference for further investigations into ocean energetics, as well as for an intended inference calculation that is more complete but also much more complex.

physics.ao-ph↗