Searcharxiv⌕ Search

arXiv subjects

Yue Wang

Publications and source records attributed to Yue Wang.

At least 613 records · Page 34Linked to original sources
arXiv2023

How to Control Hydrodynamic Force on Fluidic Pinball via Deep Reinforcement Learning

Deep reinforcement learning (DRL) for fluidic pinball, three individually rotating cylinders in the uniform flow arranged in an equilaterally triangular configuration, can learn the efficient flow control strategies due to the validity of self-learning and data-driven state estimation for complex fluid dynamic problems. In this work, we present a DRL-based real-time feedback strategy to control the hydrodynamic force on fluidic pinball, i.e., force extremum and tracking, from cylinders' rotation. By adequately designing reward functions and encoding historical observations, and after automatic learning of thousands of iterations, the DRL-based control was shown to make reasonable and valid control decisions in nonparametric control parameter space, which is comparable to and even better than the optimal policy found through lengthy brute-force searching. Subsequently, one of these results was analyzed by a machine learning model that enabled us to shed light on the basis of decision-making and physical mechanisms of the force tracking process. The finding from this work can control hydrodynamic force on the operation of fluidic pinball system and potentially pave the way for exploring efficient active flow control strategies in other complex fluid dynamic problems.

eess.SY↗
arXiv2023

Provably Efficient Offline Reinforcement Learning with Trajectory-Wise Reward

The remarkable success of reinforcement learning (RL) heavily relies on observing the reward of every visited state-action pair. In many real world applications, however, an agent can observe only a score that represents the quality of the whole trajectory, which is referred to as the {\em trajectory-wise reward}. In such a situation, it is difficult for standard RL methods to well utilize trajectory-wise reward, and large bias and variance errors can be incurred in policy evaluation. In this work, we propose a novel offline RL algorithm, called Pessimistic vAlue iteRaTion with rEward Decomposition (PARTED), which decomposes the trajectory return into per-step proxy rewards via least-squares-based reward redistribution, and then performs pessimistic value iteration based on the learned proxy reward. To ensure the value functions constructed by PARTED are always pessimistic with respect to the optimal ones, we design a new penalty term to offset the uncertainty of the proxy reward. For general episodic MDPs with large state space, we show that PARTED with overparameterized neural network function approximation achieves an $\tilde{\mathcal{O}}(D_{\text{eff}}H^2/\sqrt{N})$ suboptimality, where $H$ is the length of episode, $N$ is the total number of samples, and $D_{\text{eff}}$ is the effective dimension of the neural tangent kernel matrix. To further illustrate the result, we show that PARTED achieves an $\tilde{\mathcal{O}}(dH^3/\sqrt{N})$ suboptimality with linear MDPs, where $d$ is the feature dimension, which matches with that with neural network function approximation, when $D_{\text{eff}}=dH$. To the best of our knowledge, PARTED is the first offline RL algorithm that is provably efficient in general MDP with trajectory-wise reward.

cs.LG↗
arXiv2023

A Hyper-network Based End-to-end Visual Servoing with Arbitrary Desired Poses

Recently, several works achieve end-to-end visual servoing (VS) for robotic manipulation by replacing traditional controller with differentiable neural networks, but lose the ability to servo arbitrary desired poses. This letter proposes a differentiable architecture for arbitrary pose servoing: a hyper-network based neural controller (HPN-NC). To achieve this, HPN-NC consists of a hyper net and a low-level controller, where the hyper net learns to generate the parameters of the low-level controller and the controller uses the 2D keypoints error for control like traditional image-based visual servoing (IBVS). HPN-NC can complete 6 degree of freedom visual servoing with large initial offset. Taking advantage of the fully differentiable nature of HPN-NC, we provide a three-stage training procedure to servo real world objects. With self-supervised end-to-end training, the performance of the integrated model can be further improved in unseen scenes and the amount of manual annotations can be significantly reduced.

cs.RO↗
arXiv2023

Graph tilings in incompatibility systems

An \emph{incompatibility system} $(G,\mathcal{F})$ consists of a graph $G$ and a family $\mathcal{F}=\{F_v\}_{v\in V(G)}$ over $G$ with $F_v\subseteq \{\{e,e'\}\in {E(G)\choose 2}: e\cap e'=\{v\}\}$. We say that two edges $e,e'\in E(G)$ are \emph{incompatible} if $\{e,e'\}\in F_v$ for some $v\in V(G)$, and otherwise \emph{compatible}. A subgraph $H$ of $G$ is \emph{compatible} if every pair of edges in $H$ are compatible. An incompatibility system $(G,\mathcal{F})$ is \emph{$Δ$-bounded} if for any vertex $v$ and any edge $e$ incident with $v$, there are at most $Δ$ members of $F_v$ containing $e$. This notion was partly motivated by a concept of transition system introduced by Kotzig in 1968, and first formulated by Krivelevich, Lee and Sudakov to study the robustness of Hamiltonicity of Dirac graphs. We prove that for any $α>0$ and any graph $H$ with $h$ vertices, there exists a constant $μ>0$ such that for any sufficiently large $n$ with $n\in h\mathbb{N}$, if $G$ is an $n$-vertex graph with $δ(G)\ge(1-\frac{1}{χ^*(H)}+α)n$ and $(G,\mathcal{F})$ is a $μn$-bounded incompatibility system, then there exists a compatible $H$-factor in $G$, where the value $χ^*(H)$ is either the chromatic number $χ(H)$ or the critical chromatic number $χ_{cr}(H)$ and we provide a dichotomy as in the Kühn--Osthus result. Moreover, we give examples $H$ for which there exists an $μn$-bounded incompatibility system $(G, \mathcal{F})$ with $n\in h\mathbb{N}$ and $δ(G)\ge(1-\frac{1}{χ^*(H)}+\fracμ{2})n$ such that $G$ contains no compatible $H$-factor. Unlike in the previous work of Kühn and Osthus on embedding $H$-factors, our proof uses the lattice-based absorption method.

math.CO↗
arXiv2023

FUTR3D: A Unified Sensor Fusion Framework for 3D Detection

Sensor fusion is an essential topic in many perception systems, such as autonomous driving and robotics. Existing multi-modal 3D detection models usually involve customized designs depending on the sensor combinations or setups. In this work, we propose the first unified end-to-end sensor fusion framework for 3D detection, named FUTR3D, which can be used in (almost) any sensor configuration. FUTR3D employs a query-based Modality-Agnostic Feature Sampler (MAFS), together with a transformer decoder with a set-to-set loss for 3D detection, thus avoiding using late fusion heuristics and post-processing tricks. We validate the effectiveness of our framework on various combinations of cameras, low-resolution LiDARs, high-resolution LiDARs, and Radars. On NuScenes dataset, FUTR3D achieves better performance over specifically designed methods across different sensor combinations. Moreover, FUTR3D achieves great flexibility with different sensor configurations and enables low-cost autonomous driving. For example, only using a 4-beam LiDAR with cameras, FUTR3D (58.0 mAP) achieves on par performance with state-of-the-art 3D detection model CenterPoint (56.6 mAP) using a 32-beam LiDAR.

cs.CV↗
arXiv2023

Compressive Spectrum Sensing Using Sampling-Controlled Block Orthogonal Matching Pursuit

This paper proposes two novel schemes of wideband compressive spectrum sensing (CSS) via block orthogonal matching pursuit (BOMP) algorithm, for achieving high sensing accuracy in real time. These schemes aim to reliably recover the spectrum by adaptively adjusting the number of required measurements without inducing unnecessary sampling redundancy. To this end, the minimum number of required measurements for successful recovery is first derived in terms of its probabilistic lower bound. Then, a CSS scheme is proposed by tightening the derived lower bound, where the key is the design of a nonlinear exponential indicator through a general-purpose sampling-controlled algorithm (SCA). In particular, a sampling-controlled BOMP (SC-BOMP) is developed through a holistic integration of the existing BOMP and the proposed SCA. For fast implementation, a modified version of SC-BOMP is further developed by exploring the block orthogonality in the form of sub-coherence of measurement matrices, which allows more compressive sampling in terms of smaller lower bound of the number of measurements. Such a fast SC-BOMP scheme achieves a desired tradeoff between the complexity and the performance. Simulations demonstrate that the two SC-BOMP schemes outperform the other benchmark algorithms.

eess.SP↗
arXiv2023

How Does Imperfect Automatic Indexing Affect Semantic Search Performance?

Documents in the health domain are often annotated with semantic concepts (i.e., terms) from controlled vocabularies. As the volume of these documents gets large, the annotation work is increasingly done by algorithms. Compared to humans, automatic indexing algorithms are imperfect and may assign wrong terms to documents, which affect subsequent search tasks where queries contain these terms. In this work, we aim to understand the performance impact of using imperfectly assigned terms in Boolean semantic searches. We used MeSH terms and biomedical literature search as a case study. We implemented multiple automatic indexing algorithms on real-world Boolean queries that consist of MeSH terms, and found that (1) probabilistic logic can handle inaccurately assigned terms better than traditional Boolean logic, (2) query-level performance is mostly limited by lowest-performing terms in a query, and (3) mixing a small amount of human indexing with automatic indexing can regain excellent query-level performance. These findings provide important implications for future work on automatic indexing.

cs.IR↗
arXiv2023

Object-centric Inference for Language Conditioned Placement: A Foundation Model based Approach

We focus on the task of language-conditioned object placement, in which a robot should generate placements that satisfy all the spatial relational constraints in language instructions. Previous works based on rule-based language parsing or scene-centric visual representation have restrictions on the form of instructions and reference objects or require large amounts of training data. We propose an object-centric framework that leverages foundation models to ground the reference objects and spatial relations for placement, which is more sample efficient and generalizable. Experiments indicate that our model can achieve a 97.75% success rate of placement with only ~0.26M trainable parameters. Besides, our method generalizes better to both unseen objects and instructions. Moreover, with only 25% training data, we still outperform the top competing approach.

cs.RO↗
arXiv2023

Using Fano factors to determine certain types of gene autoregulation

The expression of one gene might be regulated by its corresponding protein, which is called autoregulation. Although gene regulation is a central topic in biology, autoregulation is much less studied. In general, it is extremely difficult to determine the existence of autoregulation with direct biochemical approaches. Nevertheless, some papers have observed that certain types of autoregulations are linked to noise levels in gene expression. We generalize these results by two propositions on discrete-state continuous-time Markov chains. These two propositions form a simple but robust method to infer the existence of autoregulation in certain scenarios from gene expression data. This method only depends on the Fano factor, namely the ratio of variance and mean of the gene expression level. Compared to other methods for inferring autoregulation, our method only requires non-interventional one-time data, and does not need to estimate parameters. Besides, our method has few restrictions on the model. We apply this method to four groups of experimental data and find some genes that might have autoregulation. Some inferred autoregulations have been verified by experiments or other theoretical works.

q-bio.MN↗
arXiv2023

UrbanGIRAFFE: Representing Urban Scenes as Compositional Generative Neural Feature Fields

Generating photorealistic images with controllable camera pose and scene contents is essential for many applications including AR/VR and simulation. Despite the fact that rapid progress has been made in 3D-aware generative models, most existing methods focus on object-centric images and are not applicable to generating urban scenes for free camera viewpoint control and scene editing. To address this challenging task, we propose UrbanGIRAFFE, which uses a coarse 3D panoptic prior, including the layout distribution of uncountable stuff and countable objects, to guide a 3D-aware generative model. Our model is compositional and controllable as it breaks down the scene into stuff, objects, and sky. Using stuff prior in the form of semantic voxel grids, we build a conditioned stuff generator that effectively incorporates the coarse semantic and geometry information. The object layout prior further allows us to learn an object generator from cluttered scenes. With proper loss functions, our approach facilitates photorealistic 3D-aware image synthesis with diverse controllability, including large camera movement, stuff editing, and object manipulation. We validate the effectiveness of our model on both synthetic and real-world datasets, including the challenging KITTI-360 dataset.

cs.CV↗
arXiv2023

Physical Backdoor Trigger Activation of Autonomous Vehicle using Reachability Analysis

Recent studies reveal that Autonomous Vehicles (AVs) can be manipulated by hidden backdoors, causing them to perform harmful actions when activated by physical triggers. However, it is still unclear how these triggers can be activated while adhering to traffic principles. Understanding this vulnerability in a dynamic traffic environment is crucial. This work addresses this gap by presenting physical trigger activation as a reachability problem of controlled dynamic system. Our technique identifies security-critical areas in traffic systems where trigger conditions for accidents can be reached, and provides intended trajectories for how those conditions can be reached. Testing on typical traffic scenarios showed the system can be successfully driven to trigger conditions with near 100% activation rate. Our method benefits from identifying AV vulnerability and enabling effective safety strategies.

cs.CR↗
arXiv2023

Optimal Smoothing Distribution Exploration for Backdoor Neutralization in Deep Learning-based Traffic Systems

Deep Reinforcement Learning (DRL) enhances the efficiency of Autonomous Vehicles (AV), but also makes them susceptible to backdoor attacks that can result in traffic congestion or collisions. Backdoor functionality is typically incorporated by contaminating training datasets with covert malicious data to maintain high precision on genuine inputs while inducing the desired (malicious) outputs for specific inputs chosen by adversaries. Current defenses against backdoors mainly focus on image classification using image-based features, which cannot be readily transferred to the regression task of DRL-based AV controllers since the inputs are continuous sensor data, i.e., the combinations of velocity and distance of AV and its surrounding vehicles. Our proposed method adds well-designed noise to the input to neutralize backdoors. The approach involves learning an optimal smoothing (noise) distribution to preserve the normal functionality of genuine inputs while neutralizing backdoors. By doing so, the resulting model is expected to be more resilient against backdoor attacks while maintaining high accuracy on genuine inputs. The effectiveness of the proposed method is verified on a simulated traffic system based on a microscopic traffic simulator, where experimental results showcase that the smoothed traffic controller can neutralize all trigger samples and maintain the performance of relieving traffic congestion

cs.LG↗
arXiv2023

DAMS-LIO: A Degeneration-Aware and Modular Sensor-Fusion LiDAR-inertial Odometry

With robots being deployed in increasingly complex environments like underground mines and planetary surfaces, the multi-sensor fusion method has gained more and more attention which is a promising solution to state estimation in the such scene. The fusion scheme is a central component of these methods. In this paper, a light-weight iEKF-based LiDAR-inertial odometry system is presented, which utilizes a degeneration-aware and modular sensor-fusion pipeline that takes both LiDAR points and relative pose from another odometry as the measurement in the update process only when degeneration is detected. Both the Cramer-Rao Lower Bound (CRLB) theory and simulation test are used to demonstrate the higher accuracy of our method compared to methods using a single observation. Furthermore, the proposed system is evaluated in perceptually challenging datasets against various state-of-the-art sensor-fusion methods. The results show that the proposed system achieves real-time and high estimation accuracy performance despite the challenging environment and poor observations.

cs.RO↗
arXiv2023

Discovery of the linear energy-dependence of the spectral lag of X-ray bursts from SGR J1935+2154

Spectral lag of the low-energy photons with respect to the high-energy ones is a common astrophysical phenomenon (such as Gamma-ray bursts and the Crab pulsar) and may serve as a key probe to the underlying radiation mechanism. However, spectral lag in keV range of the magnetar bursts has not been systematically studied yet. In this work, we perform a detailed spectral lag analysis with the Li-CCF method for SGR J1935+2154 bursts observed by {\it Insight}-HXMT, GECAM and Fermi/GBM from July 2014 to Jan 2022. We discover that the spectral lags of about 61\% (non-zero significance >1$σ$) bursts from SGR J1935+2154 are linearly dependent on the photon energy ($E$) with $t_{\rm lag}(E)=α(E/{\rm keV})+C$, which may be explained by a linear change of the temperature of the blackbody-emitting plasma with time. The distribution of the slope ($α$) approximately follows a Gaussian function with mean and standard deviation of 0.02 ms/keV (i.e. high-energy photons arrive earlier) and 0.02 ms/keV, respectively. We also find that the distribution can be well fitted with three Gaussians with mean values of $\sim$ -0.009, 0.013 and 0.039 ms/keV, which may correspond to different origins of the bursts. These spectral lag features may have important implications on the magnetar bursts.

astro-ph.HE↗
arXiv2023

GOOD: General Optimization-based Fusion for 3D Object Detection via LiDAR-Camera Object Candidates

3D object detection serves as the core basis of the perception tasks in autonomous driving. Recent years have seen the rapid progress of multi-modal fusion strategies for more robust and accurate 3D object detection. However, current researches for robust fusion are all learning-based frameworks, which demand a large amount of training data and are inconvenient to implement in new scenes. In this paper, we propose GOOD, a general optimization-based fusion framework that can achieve satisfying detection without training additional models and is available for any combinations of 2D and 3D detectors to improve the accuracy and robustness of 3D detection. First we apply the mutual-sided nearest-neighbor probability model to achieve the 3D-2D data association. Then we design an optimization pipeline that can optimize different kinds of instances separately based on the matching result. Apart from this, the 3D MOT method is also introduced to enhance the performance aided by previous frames. To the best of our knowledge, this is the first optimization-based late fusion framework for multi-modal 3D object detection which can be served as a baseline for subsequent research. Experiments on both nuScenes and KITTI datasets are carried out and the results show that GOOD outperforms by 9.1\% on mAP score compared with PointPillars and achieves competitive results with the learning-based late fusion CLOCs.

cs.CV↗
arXiv2023

Mathematical models for order of mutation problem in myeloproliferative neoplasm: non-additivity and non-commutativity

In some patients of myeloproliferative neoplasm, two genetic mutations can be found: JAK2 V617F and TET2. When one mutation is present or not, the other mutation has different effects on regulating gene expressions. Besides, when both mutations are present, the order of occurrence might make a difference. In this paper, we build nonlinear ordinary differential equation models and Markov chain models to explain such phenomena.

q-bio.PE↗
arXiv2023

Sharp Lower Bound for the Blow-up Rate of Solutions to the Magnetic Zakharov System without the Skin Effect

In this paper, we consider the Cauchy problem of the magnetic Zakharov system in two-dimensional space: \[ \begin{cases} & i E_{1t}+ΔE_1-n E_1+ηE_2 (E_1\overline{E_2}-\overline{E_1} E_2)=0, \\ & i E_{2t}+ΔE_2-n E_2+ηE_1(\overline{E_1} E_2-E_1\overline{E_2})=0, \\ & n_t+\nabla \cdot \textbf{v}=0, \\ & \textbf{v}_t+\nabla n+\nabla (|E_1|^2+|E_2|^2)=0, \\ \end{cases} \tag{G-Z} \] with initial data $\left(E_{10}(x),E_{20}(x),n_{0}(x),\mathbf{v}_{0}(x)\right)$, which describes the spontaneous generation of a magnetic field without the skin effect in a cold plasma, where $η>0$ is a physical constant coefficient. The two nonlinear terms generated by the cold magnetic field bring in a different difficulty from that for the classical Zakharov system. Assuming the initial mass satisfies the following estimates: \begin{gather*} \frac{||Q||_{L^2(\mathbb{R}^2)}^2}{1+η} <||E_{10}||_{L^2(\mathbb{R}^2)}^2+||E_{20}||_{L^2(\mathbb{R}^2)}^2 <\frac{||Q||_{L^2(\mathbb{R}^2)}^2}η, \end{gather*} where $Q$ is the unique radially positive solution of the equation $-ΔV+V=V^3 $, we prove that there is a constant $c>0$ depending only on the initial data such that for $t$ near $T$ (the blow-up time), \begin{gather*} \left\|\left(E_1,E_2,n,\textbf{v}\right)\right\|_{H^1(\mathbb{R}^2)\times H^1(\mathbb{R}^2)\times L^2(\mathbb{R}^2)\times L^2(\mathbb{R}^2)}\geqslant \frac{c}{ T-t }. \end{gather*} As the magnetic coefficient $η$ tends to $0$, the blow-up rate recovers the result for the classical 2-D Zakharov system due to Merle \cite{25Frank}. For any size positive $η$, under the current assumption on the initial mass, we give a mathematically rigorous justification for the fact that the presence of magnetic effects without the skin effect in the cold plasma does not change the optimal lower bound for the blow-up rates.

math.AP↗
arXiv2023

FreeNeRF: Improving Few-shot Neural Rendering with Free Frequency Regularization

Novel view synthesis with sparse inputs is a challenging problem for neural radiance fields (NeRF). Recent efforts alleviate this challenge by introducing external supervision, such as pre-trained models and extra depth signals, and by non-trivial patch-based rendering. In this paper, we present Frequency regularized NeRF (FreeNeRF), a surprisingly simple baseline that outperforms previous methods with minimal modifications to the plain NeRF. We analyze the key challenges in few-shot neural rendering and find that frequency plays an important role in NeRF's training. Based on the analysis, we propose two regularization terms. One is to regularize the frequency range of NeRF's inputs, while the other is to penalize the near-camera density fields. Both techniques are ``free lunches'' at no additional computational cost. We demonstrate that even with one line of code change, the original NeRF can achieve similar performance as other complicated methods in the few-shot setting. FreeNeRF achieves state-of-the-art performance across diverse datasets, including Blender, DTU, and LLFF. We hope this simple baseline will motivate a rethinking of the fundamental role of frequency in NeRF's training under the low-data regime and beyond.

cs.CV↗