arXiv subjects

Yue Wu

Publications and source records attributed to Yue Wu.

At least 289 records · Page 16Linked to original sources

arXiv2022

Optimizing Video Prediction via Video Frame Interpolation

Video prediction is an extrapolation task that predicts future frames given past frames, and video frame interpolation is an interpolation task that estimates intermediate frames between two frames. We have witnessed the tremendous advancement of video frame interpolation, but the general video prediction in the wild is still an open question. Inspired by the photo-realistic results of video frame interpolation, we present a new optimization framework for video prediction via video frame interpolation, in which we solve an extrapolation problem based on an interpolation model. Our video prediction framework is based on optimization with a pretrained differentiable video frame interpolation module without the need for a training dataset, and thus there is no domain gap issue between training and test data. Also, our approach does not need any additional information such as semantic or instance maps, which makes our framework applicable to any video. Extensive experiments on the Cityscapes, KITTI, DAVIS, Middlebury, and Vimeo90K datasets show that our video prediction results are robust in general scenarios, and our approach outperforms other video prediction methods that require a large amount of training data or extra semantic information.

cs.CV↗

arXiv2022

Geometric Policy Iteration for Markov Decision Processes

Recently discovered polyhedral structures of the value function for finite state-action discounted Markov decision processes (MDP) shed light on understanding the success of reinforcement learning. We investigate the value function polytope in greater detail and characterize the polytope boundary using a hyperplane arrangement. We further show that the value space is a union of finitely many cells of the same hyperplane arrangement and relate it to the polytope of the classical linear programming formulation for MDPs. Inspired by these geometric properties, we propose a new algorithm, Geometric Policy Iteration (GPI), to solve discounted MDPs. GPI updates the policy of a single state by switching to an action that is mapped to the boundary of the value function polytope, followed by an immediate update of the value function. This new update rule aims at a faster value improvement without compromising computational efficiency. Moreover, our algorithm allows asynchronous updates of state values which is more flexible and advantageous compared to traditional policy iteration when the state set is large. We prove that the complexity of GPI achieves the best known bound $\mathcal{O}\left(\frac{|\mathcal{A}|}{1 - γ}\log \frac{1}{1-γ}\right)$ of policy iteration and empirically demonstrate the strength of GPI on MDPs of various sizes.

cs.LG↗

arXiv2022

The backward Euler-Maruyama method for invariant measures of stochastic differential equations with super-linear coefficients

The backward Euler-Maruyama (BEM) method is employed to approximate the invariant measure of stochastic differential equations, where both the drift and the diffusion coefficient are allowed to grow super-linearly. The existence and uniqueness of the invariant measure of the numerical solution generated by the BEM method are proved and the convergence of the numerical invariant measure to the underlying one is shown. Simulations are provided to illustrate the theoretical results and demonstrate the application of our results in the area of system control.

math.PR↗

arXiv2022

Riesz transform associated with the fractional Fourier transform and applications in image edge detection

The fractional Hilbert transform was introduced by Zayed [30, Zayed, 1998] and has been widely used in signal processing. In view of is connection with the fractional Fourier transform, Chen, the first, second and fourth authors of this paper in [6, Chen et al., 2021] studied the fractional Hilbert transform and other fractional multiplier operators on the real line. The present paper is concerned with a natural extension of the fractional Hilbert transform to higher dimensions: this extension is the fractional Riesz transform which is defined by multiplication which a suitable chirp function on the fractional Fourier transform side. In addition to a thorough study of the fractional Riesz transforms, in this work we also investigate the boundedness of singular integral operators with chirp functions on rotation invariant spaces, chirp Hardy spaces and their relation to chirp BMO spaces, as well as applications of the theory of fractional multipliers in partial differential equations. Through numerical simulation, we provide physical and geometric interpretations of high-dimensional fractional multipliers. Finally, we present an application of the fractional Riesz transforms in edge detection which verifies a hypothesis insinuated in [26, Xu et al., 2016]. In fact our numerical implementation confirms that amplitude, phase, and direction information can be simultaneously extracted by controlling the order of the fractional Riesz transform.

math.FA↗

arXiv2022

The Galerkin analysis for the random periodic solution of semilinear stochastic evolution equations

In this paper we study the numerical method for approximating the random periodic solution of semiliear stochastic evolution equations. The main challenge lies in proving a convergence over an infinite time horizon while simulating infinite-dimensional objects. We first show the existence and uniqueness of the random periodic solution to the equation as the limit of the pull-back flows of the equation, and observe that its mild form is well-defined in the intersection of a family of decreasing Hilbert spaces. Then we propose a Galerkin-type exponential integrator scheme and establish its convergence rate of the strong error to the mild solution, where the order of convergence directly depends on the space (among the family of Hilbert spaces) for the initial point to live. We finally conclude with the best order of convergence that is arbitrarily close to 0.5.

math.PR↗

arXiv2022

Nearly Minimax Optimal Regret for Learning Infinite-horizon Average-reward MDPs with Linear Function Approximation

We study reinforcement learning in an infinite-horizon average-reward setting with linear function approximation, where the transition probability function of the underlying Markov Decision Process (MDP) admits a linear form over a feature mapping of the current state, action, and next state. We propose a new algorithm UCRL2-VTR, which can be seen as an extension of the UCRL2 algorithm with linear function approximation. We show that UCRL2-VTR with Bernstein-type bonus can achieve a regret of $\tilde{O}(d\sqrt{DT})$, where $d$ is the dimension of the feature mapping, $T$ is the horizon, and $\sqrt{D}$ is the diameter of the MDP. We also prove a matching lower bound $\tildeΩ(d\sqrt{DT})$, which suggests that the proposed UCRL2-VTR is minimax optimal up to logarithmic factors. To the best of our knowledge, our algorithm is the first nearly minimax optimal RL algorithm with function approximation in the infinite-horizon average-reward setting.

cs.LG↗

arXiv2022

Lesion Localization in OCT by Semi-Supervised Object Detection

Over 300 million people worldwide are affected by various retinal diseases. By noninvasive Optical Coherence Tomography (OCT) scans, a number of abnormal structural changes in the retina, namely retinal lesions, can be identified. Automated lesion localization in OCT is thus important for detecting retinal diseases at their early stage. To conquer the lack of manual annotation for deep supervised learning, this paper presents a first study on utilizing semi-supervised object detection (SSOD) for lesion localization in OCT images. To that end, we develop a taxonomy to provide a unified and structured viewpoint of the current SSOD methods, and consequently identify key modules in these methods. To evaluate the influence of these modules in the new task, we build OCT-SS, a new dataset consisting of over 1k expert-labeled OCT B-scan images and over 13k unlabeled B-scans. Extensive experiments on OCT-SS identify Unbiased Teacher (UnT) as the best current SSOD method for lesion localization. Moreover, we improve over this strong baseline, with mAP increased from 49.34 to 50.86.

cs.CV↗

arXiv2022

Stellar Atmospheric Parameters of M-type Stars from LAMOST DR8

The Large Sky Area Multi-Object Fiber Spectroscopic Telescope (LAMOST) Low Resolution Spectroscopic Survey (LRS) provides massive spectroscopic data of M-type stars, and the derived stellar parameters could bring vital help to various studies. We adopt the ULySS package to perform $χ^2$ minimization with model spectra generated from the MILES interpolator, and determine the stellar atmospheric parameters for the M-type stars from LAMOST LRS Data Release (DR) 8. Comparison with the stellar parameters from APOGEE Stellar Parameter and Chemical Abundance Pipeline (ASPCAP) suggests that most of our results have good consistency. For M dwarfs, we achieve dispersions better than 74 K, 0.19 dex and 0.16 dex for $T_{\rm eff}$, $\log{g}$ and [Fe/H], while for M giants, the internal uncertainties are 58 K, 0.32 dex and 0.26 dex, respectively. Compared to ASPCAP we also find a systematic underestimation of $Δ{T_{\rm eff}} =$ $-$176 K for M dwarfs, and a systematic overestimation of $Δ{\log{g}} =$ 0.30 dex for M giants. However, such differences are less significant when we make comparison with common stars from other literature, which indicates that systematic biases exist in the difference of ASPCAP and other measurements. A catalog of 763,136 spectra corresponding to 616,314 M-type stars with derived stellar parameters is presented. We determine the stellar parameters for stars with $T_{\rm eff}$ higher than 2,900 K, with $\log{g}$ from -0.24 dex to 5.9 dex. The typical precisions are 45 K, 0.25 dex and 0.22 dex, for $T_{\rm eff}$, $\log{g}$ and [Fe/H], respectively, which are estimated from the duplicate observations of the same stars.

astro-ph.SR↗

arXiv2022

Game Theoretic Consequences of Resident Matching

The resident matching algorithm, Gale-Shapley, currently used by SF Match and the National Residency Match Program (NRMP), has been in use for over 50 years without fundamental alteration. The algorithm is a 'stable-marriage' method that favors applicant outcomes. However, in these 50 years, there has been a big shift in the supply and demand of applicants and programs. These changes along with the way the Match is implemented have induced a costly race among applicants to apply and interview at as many programs as possible. Meanwhile programs also incur high costs as they maximize their probability of matching by interviewing as many candidates as possible.

econ.TH↗

arXiv2022

Real-time Online Multi-Object Tracking in Compressed Domain

Recent online Multi-Object Tracking (MOT) methods have achieved desirable tracking performance. However, the tracking speed of most existing methods is rather slow. Inspired from the fact that the adjacent frames are highly relevant and redundant, we divide the frames into key and non-key frames respectively and track objects in the compressed domain. For the key frames, the RGB images are restored for detection and data association. To make data association more reliable, an appearance Convolutional Neural Network (CNN) which can be jointly trained with the detector is proposed. For the non-key frames, the objects are directly propagated by a tracking CNN based on the motion information provided in the compressed domain. Compared with the state-of-the-art online MOT methods,our tracker is about 6x faster while maintaining a comparable tracking performance.

cs.CV↗

arXiv2022

DoF of a Cooperative X-Channel with an Application to Distributed Computing

We consider a cooperative X-channel with $\sf K$ transmitters (TXs) and $\sf K$ receivers (Rxs) where Txs and Rxs are gathered into groups of size $\sf r$ respectively. Txs belonging to the same group cooperate to jointly transmit a message to each of the $\sf K- \sf r$ Rxs in all other groups, and each Rx individually decodes all its intended messages. By introducing a new interference alignment (IA) scheme, we prove that when $\sf K/\sf r$ is an integer the sum Degrees of Freedom (SDoF) of this channel is lower bounded by $2\sf r$ if $\sf K/\sf r \in \{2,3\}$ and by $\frac{\sf K(\sf K-\sf r)-\sf r^2}{2\sf K-3\sf r}$ if $\sf K/\sf r \geq 4$. We also prove that the SDoF is upper bounded by $\frac{\sf K(\sf K-\sf r)}{2\sf K-3\sf r}$. The proposed IA scheme finds application in a wireless distributed MapReduce framework, where it improves the normalized data delivery time (NDT) compared to the state of the art.

cs.IT↗

arXiv2022

A direct imaging method for the exterior and interior inverse scattering problems

This paper is concerned with the inverse acoustic scattering problems by an obstacle or a cavity with a sound-soft or a sound-hard boundary. A direct imaging method relying on the boundary conditions is proposed for reconstructing the shape of the obstacle or cavity. First, the scattered fields are approximated by the Fourier-Bessel functions with the measurements on a closed curve. Then, the indicator functions are established by the superpositions of the total fields or their derivatives to the incident point sources. We prove that the indicator functions vanish only on the boundary of the obstacle or cavity. Numerical examples are also included to demonstrate the effectiveness of the method.

math.NA↗

arXiv2022

The China Trade Shock and the ESG Performances of US firms

How does import competition from China affect engagement on ESG initiatives by US corporates? On the one hand, reduced profitability due to import competition and lagging ESG performance of Chinese exporters can disincentivize US firms to put more resources to ESG initiatives. On the other hand, the shift from labor-intensive production to capital/technology-intensive production along with offshoring may improve the US company's ESG performance. Moreover, US companies have incentives to actively pursue more ESG engagement to differentiate from Chinese imports. Exploiting a trade policy in which US congress granted China the Permanent Normal Trade Relations and the resulting change in expected tariff rates on Chinese imports, we find that greater import competition from China leads to an increase in the US company's ESG performance. The improvement primarily stems from "doing more positives" and from more involvement on environmental initiatives. Indirect and direct evidence shows that the improvement is not driven by the change in production process or offshoring, but is consistent with product differentiation. Our results suggest that the trade shock from China has significant impact on the US company's ESG performance.

q-fin.GN↗

arXiv2022

Erasure conversion for fault-tolerant quantum computing in alkaline earth Rydberg atom arrays

Executing quantum algorithms on error-corrected logical qubits is a critical step for scalable quantum computing, but the requisite numbers of qubits and physical error rates are demanding for current experimental hardware. Recently, the development of error correcting codes tailored to particular physical noise models has helped relax these requirements. In this work, we propose a qubit encoding and gate protocol for ${}^{171}$Yb neutral atom qubits that converts the dominant physical errors into erasures, that is, errors in known locations. The key idea is to encode qubits in a metastable electronic level, such that gate errors predominantly result in transitions to disjoint subspaces whose populations can be continuously monitored via fluorescence. We estimate that 98% of errors can be converted into erasures. We quantify the benefit of this approach via circuit-level simulations of the surface code, finding a threshold increase from 0.937% to 4.15%. We also observe a larger code distance near the threshold, leading to a faster decrease in the logical error rate for the same number of physical qubits, which is important for near-term implementations. Erasure conversion should benefit any error correcting code, and may also be applied to design new gates and encodings in other qubit platforms.

quant-ph↗

arXiv2021

Spatially Constrained GAN for Face and Fashion Synthesis

Image generation has raised tremendous attention in both academic and industrial areas, especially for the conditional and target-oriented image generation, such as criminal portrait and fashion design. Although the current studies have achieved preliminary results along this direction, they always focus on class labels as the condition where spatial contents are randomly generated from latent vectors. Edge details are usually blurred since spatial information is difficult to preserve. In light of this, we propose a novel Spatially Constrained Generative Adversarial Network (SCGAN), which decouples the spatial constraints from the latent vector and makes these constraints feasible as additional controllable signals. To enhance the spatial controllability, a generator network is specially designed to take a semantic segmentation, a latent vector and an attribute-level label as inputs step by step. Besides, a segmentor network is constructed to impose spatial constraints on the generator. Experimentally, we provide both visual and quantitative results on CelebA and DeepFashion datasets, and demonstrate that the proposed SCGAN is very effective in controlling the spatial contents as well as generating high-quality images.

cs.CV↗

arXiv2021

Unsupervised cross domain learning with applications to 7 layer segmentation of OCTs

Unsupervised cross domain adaptation for OCT 7 layer segmentation and other medical applications where labeled training data is only available in a source domain and unavailable in the target domain. Our proposed method helps generalize of deep learning to many areas in the medical field where labeled training data are expensive and time consuming to acquire or where target domains are too novel to have had labelling.

eess.IV↗

arXiv2021

Signal Encryption Strategy based on Domain change of the Fractional Fourier Transform

This paper provides a double encryption algorithm that uses the lack of invertibility of the fractional Fourier transform (FRFT) on $L^{1}$. One encryption key is a function, which maps a ``good" $L^{2}$-signal to a ``bad" $L^{1}$-signal. The FRFT parameter which describes the rotation associated with this operator on the time-frequency plane provides the other encryption key. With the help of approximate identities, such as of the Abel and Gauss means of the FRFT established in \cite{CFGW}, we recover the encrypted signal on the FRFT domain. This design of an encryption algorithm seems new even when using the classical Fourier transform. Finally, the feasibility of the new strategy is verified by simulation and audio examples.

cs.IT↗

arXiv2021

ADDS: Adaptive Differentiable Sampling for Robust Multi-Party Learning

Distributed multi-party learning provides an effective approach for training a joint model with scattered data under legal and practical constraints. However, due to the quagmire of a skewed distribution of data labels across participants and the computation bottleneck of local devices, how to build smaller customized models for clients in various scenarios while providing updates appliable to the central model remains a challenge. In this paper, we propose a novel adaptive differentiable sampling framework (ADDS) for robust and communication-efficient multi-party learning. Inspired by the idea of dropout in neural networks, we introduce a network sampling strategy in the multi-party setting, which distributes different subnets of the central model to clients for updating, and the differentiable sampling rates allow each client to extract optimal local architecture from the supernet according to its private data distribution. The approach requires minimal modifications to the existing multi-party learning structure, and it is capable of integrating local updates of all subnets into the supernet, improving the robustness of the central model. The proposed framework significantly reduces local computation and communication costs while speeding up the central model convergence, as we demonstrated through experiments on real-world datasets.

cs.LG↗