arXiv subjects

Chenguang Liu

Publications and source records attributed to Chenguang Liu.

33 records · Page 2Linked to original sources

arXiv2024

YOLC: You Only Look Clusters for Tiny Object Detection in Aerial Images

Detecting objects from aerial images poses significant challenges due to the following factors: 1) Aerial images typically have very large sizes, generally with millions or even hundreds of millions of pixels, while computational resources are limited. 2) Small object size leads to insufficient information for effective detection. 3) Non-uniform object distribution leads to computational resource wastage. To address these issues, we propose YOLC (You Only Look Clusters), an efficient and effective framework that builds on an anchor-free object detector, CenterNet. To overcome the challenges posed by large-scale images and non-uniform object distribution, we introduce a Local Scale Module (LSM) that adaptively searches cluster regions for zooming in for accurate detection. Additionally, we modify the regression loss using Gaussian Wasserstein distance (GWD) to obtain high-quality bounding boxes. Deformable convolution and refinement methods are employed in the detection head to enhance the detection of small objects. We perform extensive experiments on two aerial image datasets, including Visdrone2019 and UAVDT, to demonstrate the effectiveness and superiority of our proposed approach. Code is available at https://github.com/dawn-ech/YOLC.

cs.CV↗

arXiv2024

Subsampling Error in Stochastic Gradient Langevin Diffusions

The Stochastic Gradient Langevin Dynamics (SGLD) are popularly used to approximate Bayesian posterior distributions in statistical learning procedures with large-scale data. As opposed to many usual Markov chain Monte Carlo (MCMC) algorithms, SGLD is not stationary with respect to the posterior distribution; two sources of error appear: The first error is introduced by an Euler--Maruyama discretisation of a Langevin diffusion process, the second error comes from the data subsampling that enables its use in large-scale data settings. In this work, we consider an idealised version of SGLD to analyse the method's pure subsampling error that we then see as a best-case error for diffusion-based subsampling MCMC methods. Indeed, we introduce and study the Stochastic Gradient Langevin Diffusion (SGLDiff), a continuous-time Markov process that follows the Langevin diffusion corresponding to a data subset and switches this data subset after exponential waiting times. There, we show the exponential ergodicity of SLGDiff and that the Wasserstein distance between the posterior and the limiting distribution of SGLDiff is bounded above by a fractional power of the mean waiting time. We bring our results into context with other analyses of SGLD.

stat.ML↗

arXiv2024

Convergence rates for Backward SDEs driven by Lévy processes

We consider Lévy processes that are approximated by compound Poisson processes and, correspondingly, BSDEs driven by Lévy processes that are approximated by BSDEs driven by their compound Poisson approximations. We are interested in the rate of convergence of the approximate BSDEs to the ones driven by the Lévy processes. The rate of convergence of the Lévy processes depends on the Blumenthal--Getoor index of the process. We derive the rate of convergence for the BSDEs in the $\mathbb L^2$-norm and in the Wasserstein distance, and show that, in both cases, this equals the rate of convergence of the corresponding Lévy process, and thus is optimal.

math.PR↗

arXiv2024

A high-order multi-time-step scheme for bond-based peridynamics

A high-order multi-time-step (MTS) scheme for the bond-based peridynamic (PD) model, an extension of classical continuous mechanics widely used for analyzing discontinuous problems like cracks, is proposed. The MTS scheme discretizes the spatial domain with a meshfree method and advances in time with a high-order Runge-Kutta method. To effectively handle discontinuities (cracks) that appear in a local subdomain in the solution, the scheme employs the Taylor expansion and Lagrange interpolation polynomials with a finer time step size, that is, coarse and fine time step sizes for smooth and discontinuous subdomains, respectively, to achieve accurate and efficient simulations. By eliminating unnecessary fine-scale resolution imposed on the entire domain, the MTS scheme outperforms the standard PD scheme by significantly reducing computational costs, particularly for problems with discontinuous solutions, as demonstrated by comprehensive theoretical analysis and numerical experiments.

math.NA↗

arXiv2023

Self-supervised Adaptive Weighting for Cooperative Perception in V2V Communications

Perception of the driving environment is critical for collision avoidance and route planning to ensure driving safety. Cooperative perception has been widely studied as an effective approach to addressing the shortcomings of single-vehicle perception. However, the practical limitations of vehicle-to-vehicle (V2V) communications have not been adequately investigated. In particular, current cooperative fusion models rely on supervised models and do not address dynamic performance degradation caused by arbitrary channel impairments. In this paper, a self-supervised adaptive weighting model is proposed for intermediate fusion to mitigate the adverse effects of channel distortion. The performance of cooperative perception is investigated in different system settings. Rician fading and imperfect channel state information (CSI) are also considered. Numerical results demonstrate that the proposed adaptive weighting algorithm significantly outperforms the benchmarks without weighting. Visualization examples validate that the proposed weighting algorithm can flexibly adapt to various channel conditions. Moreover, the adaptive weighting algorithm demonstrates good generalization to untrained channels and test datasets from different domains.

cs.CV↗

arXiv2023

Knowledge Distillation Based Semantic Communications For Multiple Users

Deep learning (DL) has shown great potential in revolutionizing the traditional communications system. Many applications in communications have adopted DL techniques due to their powerful representation ability. However, the learning-based methods can be dependent on the training dataset and perform worse on unseen interference due to limited model generalizability and complexity. In this paper, we consider the semantic communication (SemCom) system with multiple users, where there is a limited number of training samples and unexpected interference. To improve the model generalization ability and reduce the model size, we propose a knowledge distillation (KD) based system where Transformer based encoder-decoder is implemented as the semantic encoder-decoder and fully connected neural networks are implemented as the channel encoder-decoder. Specifically, four types of knowledge transfer and model compression are analyzed. Important system and model parameters are considered, including the level of noise and interference, the number of interfering users and the size of the encoder and decoder. Numerical results demonstrate that KD significantly improves the robustness and the generalization ability when applied to unexpected interference, and it reduces the performance loss when compressing the model size.

eess.SP↗

arXiv2023

Cooperative Perception with Learning-Based V2V communications

Cooperative perception has been widely used in autonomous driving to alleviate the inherent limitation of single automated vehicle perception. To enable cooperation, vehicle-to-vehicle (V2V) communication plays an indispensable role. This work analyzes the performance of cooperative perception accounting for communications channel impairments. Different fusion methods and channel impairments are evaluated. A new late fusion scheme is proposed to leverage the robustness of intermediate features. In order to compress the data size incurred by cooperation, a convolution neural network-based autoencoder is adopted. Numerical results demonstrate that intermediate fusion is more robust to channel impairments than early fusion and late fusion, when the SNR is greater than 0 dB. Also, the proposed fusion scheme outperforms the conventional late fusion using detection outputs, and autoencoder provides a good compromise between detection accuracy and bandwidth usage.

eess.SP↗

arXiv2023

A fast computational framework for the linear bond-based peridynamic model

Peridynamic (PD) theory is significant and promising in engineering and materials science; however, it imposes challenges owing to the enormous computational cost caused by its nonlocality. Our main contribution, which overcomes the restrictions of the existing fast method, is a general computational framework for the linear bond-based peridynamic models based on the meshfree method, called the matrix-structure-based fast method (MSBFM), which is suitable for the general case, including 2D/3D problems, and static/dynamic issues, as well as problems with general boundary conditions, in particular, problems with crack propagation. Consequently, we provide a general calculation flow chart. The proposed computational framework is practical and easily embedded into the existing computational algorithm. With this framework, the computational cost is reduced from $O(N^2)$ to $O(N\log N)$, and the storage request is reduced from $O(N^2)$ to $O(N)$, where N is the degree of freedom. Finally, the vast reduction of the computational and memory requirement is verified by numerical examples.

math.NA↗

arXiv2021

A Continuous-time Stochastic Gradient Descent Method for Continuous Data

Optimization problems with continuous data appear in, e.g., robust machine learning, functional data analysis, and variational inference. Here, the target function is given as an integral over a family of (continuously) indexed target functions - integrated with respect to a probability measure. Such problems can often be solved by stochastic optimization methods: performing optimization steps with respect to the indexed target function with randomly switched indices. In this work, we study a continuous-time variant of the stochastic gradient descent algorithm for optimization problems with continuous data. This so-called stochastic gradient process consists in a gradient flow minimizing an indexed target function that is coupled with a continuous-time index process determining the index. Index processes are, e.g., reflected diffusions, pure jump processes, or other Lévy processes on compact spaces. Thus, we study multiple sampling patterns for the continuous data space and allow for data simulated or streamed at runtime of the algorithm. We analyze the approximation properties of the stochastic gradient process and study its longtime behavior and ergodicity under constant and decreasing learning rates. We end with illustrating the applicability of the stochastic gradient process in a polynomial regression problem with noisy functional data, as well as in a physics-informed neural network.

cs.LG↗

arXiv2020

Rate of convergence of the Kac-like particle system

In this paper, we consider the Kac stochastic particle system associated to the spatially homogeneous Boltzmann equation for true hard potentials. We establish a rate of propagation of chaos of the particle system to the unique solution of the Boltzmann equation. We use a probabilistic coupling method and give, under suitable assumptions on the initial condition, a rate of convergence of the empirical measure of the particle system to the solution of the Boltzmann equation for this singular interaction.

math.PR↗

arXiv2020

Progressive Defense Against Adversarial Attacks for Deep Learning as a Service in Internet of Things

Nowadays, Deep Learning as a service can be deployed in Internet of Things (IoT) to provide smart services and sensor data processing. However, recent research has revealed that some Deep Neural Networks (DNN) can be easily misled by adding relatively small but adversarial perturbations to the input (e.g., pixel mutation in input images). One challenge in defending DNN against these attacks is to efficiently identifying and filtering out the adversarial pixels. The state-of-the-art defense strategies with good robustness often require additional model training for specific attacks. To reduce the computational cost without loss of generality, we present a defense strategy called a progressive defense against adversarial attacks (PDAAA) for efficiently and effectively filtering out the adversarial pixel mutations, which could mislead the neural network towards erroneous outputs, without a-priori knowledge about the attack type. We evaluated our progressive defense strategy against various attack methods on two well-known datasets. The result shows it outperforms the state-of-the-art while reducing the cost of model training by 50% on average.

cs.CR↗

arXiv2020

Statistical inference for a partially observed interacting system of Hawkes processes

We observe the actions of a $K$ sub-sample of $N$ individuals up to time $t$ for some large $K 0$ and on the sum of some function $ϕ$ of the ages of the actions of the individuals which influence him. The function $ϕ$ is unknown but we assume it rapidly decays. The aim of this paper is to estimate the parameter $p$ asymptotically as $N\to \infty$, $K\to \infty$, and $t\to \infty$. Let $m_t$ be the average number of actions per individual up to time $t$. In the subcritical case, where $m_t$ increases linearly, we build an estimator of $p$ with the rate of convergence $\frac{1}{\sqrt{K}}+\frac{N}{m_t\sqrt{K}}+\frac{N}{K\sqrt{m_t}}$. In the supercritical case, where $m_{t}$ increases exponentially fast, we build an estimator of $p$ with the rate of convergence $\frac{1}{\sqrt{K}}+\frac{N}{m_{t}\sqrt{K}}$.

math.ST↗

arXiv2020

rIoT: Enabling Seamless Context-Aware Automation in the Internet of Things

Advances in mobile computing capabilities and an increasing number of Internet of Things (IoT) devices have enriched the possibilities of the IoT but have also increased the cognitive load required of IoT users. Existing context-aware systems provide various levels of automation in the IoT. Many of these systems adaptively take decisions on how to provide services based on assumptions made a priori. The approaches are difficult to personalize to an individual's dynamic environment, and thus today's smart IoT spaces often demand complex and specialized interactions with the user in order to provide tailored services. We propose rIoT, a framework for seamless and personalized automation of human-device interaction in the IoT. rIoT leverages existing technologies to operate across heterogeneous devices and networks to provide a one-stop solution for device interaction in the IoT. We show how rIoT exploits similarities between contexts and employs a decision-tree like method to adaptively capture a user's preferences from a small number of interactions with the IoT space. We measure the performance of rIoT on two real-world data sets and a real mobile device in terms of accuracy, learning speed, and latency in comparison to two state-of-the-art machine learning algorithms.

cs.CY↗

arXiv2019

Faster super-resolution imaging with auto-correlation two-step deconvolution

Despite super-resolution fluorescence blinking microscopes break the diffraction limit, the intense phototoxic illumination and long-term image sequences thus far still pose to major challenges in visualizing live-organisms. Here, we proposed a super-resolution method based on auto-correlation two-step deconvolution (SACD) to enhance the temporal resolution at lower signal intensity levels. Unlike conventional techniques, such as super-resolution optical fluctuation imaging (SOFI) or stochastic optical reconstruction microscopy (STORM), our model allows 16 frames to generate super-resolution images, without noticeable degradation in recording quality. We demonstrate SACD both in simulated predictions and experimental validations, with the resulting spatial resolution of 64 nm and 2-10 fold speed improvements. The use of low signal-to-noise ratio acquisition of image sequences, our SACD enables fast, background-free, super-resolution imaging that maybe advance implemented as a suitable tool for rapid live-cells imaging.

physics.optics↗

arXiv2019

Central limit theorem for a partially observed interacting system of Hawkes processes

We observe the actions of a $K$ sub-sample of $N$ individuals up to time $t$ for some large $K\le N$. We model the relationships of individuals by i.i.d. Bernoulli($p$)-random variables, where $p\in (0,1]$ is an unknown parameter. The rate of action of each individual depends on some unknown parameter $μ> 0$ and on the sum of some function $ϕ$ of the ages of the actions of the individuals which influence him. The parameters $μ$ and $ϕ$ are considered as nuisance parameters. The aim of this paper is to obtain a central limit theorem for the estimator of $p$ that we introduced in \cite{D}, both in the subcritical and supercritical cases.

math.ST↗