arXiv subjects

Yu Zhang

Publications and source records attributed to Yu Zhang.

At least 775 records · Page 43Linked to original sources

arXiv2024

Enhancing Sharpness-Aware Minimization by Learning Perturbation Radius

Sharpness-aware minimization (SAM) is to improve model generalization by searching for flat minima in the loss landscape. The SAM update consists of one step for computing the perturbation and the other for computing the update gradient. Within the two steps, the choice of the perturbation radius is crucial to the performance of SAM, but finding an appropriate perturbation radius is challenging. In this paper, we propose a bilevel optimization framework called LEarning the perTurbation radiuS (LETS) to learn the perturbation radius for sharpness-aware minimization algorithms. Specifically, in the proposed LETS method, the upper-level problem aims at seeking a good perturbation radius by minimizing the squared generalization gap between the training and validation losses, while the lower-level problem is the SAM optimization problem. Moreover, the LETS method can be combined with any variant of SAM. Experimental results on various architectures and benchmark datasets in computer vision and natural language processing demonstrate the effectiveness of the proposed LETS method in improving the performance of SAM.

cs.LG↗

arXiv2024

Saturation Numbers for Linear Forests $P_7+tP_2$

Let $H$ be a fixed graph, a graph G is $H$-saturated if it has no copy of $H$ in $G$, but the addition of any edge in $E(\overline G)$ to $G$ results in an $H$-subgraph. The saturation number sat$(n,H)$ is the minimum number of edges in an $H$-saturated graph on $n$ vertices. In this paper, we determine the saturation number sat$(n,P_7+tP_2)$ for $n\geq \frac {14}{5}t+27$ and characterize the extremal graphs for $n\geq \frac{14}{13}(3t+25)$.

math.CO↗

arXiv2024

Building Decision Making Models Through Language Model Regime

We propose a novel approach for decision making problems leveraging the generalization capabilities of large language models (LLMs). Traditional methods such as expert systems, planning algorithms, and reinforcement learning often exhibit limited generalization, typically requiring the training of new models for each unique task. In contrast, LLMs demonstrate remarkable success in generalizing across varied language tasks, inspiring a new strategy for training decision making models. Our approach, referred to as "Learning then Using" (LTU), entails a two-stage process. Initially, the \textit{learning} phase develops a robust foundational decision making model by integrating diverse knowledge from various domains and decision making contexts. The subsequent \textit{using} phase refines this foundation model for specific decision making scenarios. Distinct from other studies that employ LLMs for decision making through supervised learning, our LTU method embraces a versatile training methodology that combines broad pre-training with targeted fine-tuning. Experiments in e-commerce domains such as advertising and search optimization have shown that LTU approach outperforms traditional supervised learning regimes in decision making capabilities and generalization. The LTU approach is the first practical training architecture for both single-step and multi-step decision making tasks combined with LLMs, which can be applied beyond game and robot domains. It provides a robust and adaptable framework for decision making, enhances the effectiveness and flexibility of various systems in tackling various challenges.

cs.CL↗

arXiv2024

Enhanced Prediction of Multi-Agent Trajectories via Control Inference and State-Space Dynamics

In the field of autonomous systems, accurately predicting the trajectories of nearby vehicles and pedestrians is crucial for ensuring both safety and operational efficiency. This paper introduces a novel methodology for trajectory forecasting based on state-space dynamic system modeling, which endows agents with models that have tangible physical implications. To enhance the precision of state estimations within the dynamic system, the paper also presents a novel modeling technique for control variables. This technique utilizes a newly introduced model, termed "Mixed Mamba," to derive initial control states, thereby improving the predictive accuracy of these variables. Moverover, the proposed approach ingeniously integrates graph neural networks with state-space models, effectively capturing the complexities of multi-agent interactions. This combination provides a robust and scalable framework for forecasting multi-agent trajectories across a range of scenarios. Comprehensive evaluations demonstrate that this model outperforms several established benchmarks across various metrics and datasets, highlighting its significant potential to advance trajectory forecasting in autonomous systems.

cs.RO↗

arXiv2024

Shadow of Kerr black hole surrounded by a cloud of strings in Rastall gravity and constraints from M87*

Motivated by the first image of a black hole captured by the Event Horizon Telescope (EHT), there has been a surge of research using observations of black hole shadows to test theories of gravity. In this paper, we carry out a study related to the shadow of Kerr black holes surrounded by a cloud of strings in Rastall gravity, which deviates from the Kerr black hole due to the presence of the string parameter $a_0$ and the parameter $β$. The horizons, ergospheres, and photon region of the black hole are shown. Moreover, we explore the shadow and observations of the black hole, which are closely linked to the parameters $a_0$ and $β$. By treating \text{M87*} as a Kerr black hole surrounded by a cloud of strings under Rastall gravity, we constrain the black hole parameters using the EHT observations. For a given $β$, the circularity deviation of the black hole obeys $ΔC\lesssim0.1$ in all regions. The angular diameter $θ_{d}=42\pm3μas$ provides the upper bound of parameters $a$ and $a_0$ for fixed $β$. The shadow axis ratio satisfies the observation data of EHT ($1<D_x\lesssim4/3$) in the whole space for a given $β$. These results are consistent with the public information from EHT. In other words, candidates for real astrophysical black holes can be Kerr black holes surrounded by a cloud of strings in Rastall gravity.

gr-qc↗

arXiv2024

Quantum Lotka-Volterra dynamics

Physical systems that display competitive non-linear dynamics have played a key role in the development of mathematical models of Nature. Important examples include predator-prey models in ecology, biology, consumer-resource models in economics, and reaction-diffusion equations in chemical reactions. However, as real world systems are embedded in complex environments, where it is difficult or even impossible to control external parameters, quantitative comparison between measurements and simple models remains challenging. This motivates the search for competitive dynamics in isolated physical systems, with precise control. An ideal candidate is laser excitation in dilute atomic ensembles. For example, atoms in highly-excited Rydberg states display rich many-body dynamics including ergodicity breaking, synchronisation and time crystals. Here, we demonstrate predator-prey dynamics by laser excitation and ionisation of Rydberg atoms in a room temperature vapour cell. Ionisation of excited atoms produce electric fields that suppress further excitation. This starves the ionisation process of resource, giving rise to predator-prey dynamics. By comparing our results to the Lotka-Volterra model, we demonstrate that as well applications in non-linear dynamics, our experiment has applications in metrology, and remote sensing of localised plasmas.

quant-ph↗

arXiv2024

Quasinormal modes, Hawking radiation and absorption of the massless scalar field for Bardeen black hole surrounded by perfect fluid dark matter

We study the quasinormal modes, Hawking radiation and absorption cross-section of the Bardeen black hole surrounded by perfect fluid dark matter for a massless scalar field. Our results show that the oscillation frequency of quasinormal modes is enhanced as magnetic charge $g$ or the dark matter parameter $α$ increases. For damping rate of quasinormal modes, the influence of them is different. Specifically, the increase of dark matter parameter $α$ makes the damping rate increasing at first and then decreasing. While the damping rate is continuously decreasing with the increase of the magnetic charge $g$. Moreover, we find that the increase of the dark matter parameter $α$ enhances the power emission spectrum whereas magnetic charge $g$ suppresses it. This means that the lifespan of black holes increases for smaller value of $α$ and larger value of $g$ when other parameters are fixed. Finally, the absorption cross-section of the considered black hole is calculated with the help of the partial wave approach. Our results suggest that the absorption cross-section decreases with the dark matter parameter $α$ or the magnetic charge $g$ increasing.

physics.gen-ph↗

arXiv2024

Lite-Mind: Towards Efficient and Robust Brain Representation Network

The limited data availability and the low signal-to-noise ratio of fMRI signals lead to the challenging task of fMRI-to-image retrieval. State-of-the-art MindEye remarkably improves fMRI-to-image retrieval performance by leveraging a large model, i.e., a 996M MLP Backbone per subject, to align fMRI embeddings to the final hidden layer of CLIP's Vision Transformer (ViT). However, significant individual variations exist among subjects, even under identical experimental setups, mandating the training of large subject-specific models. The substantial parameters pose significant challenges in deploying fMRI decoding on practical devices. To this end, we propose Lite-Mind, a lightweight, efficient, and robust brain representation learning paradigm based on Discrete Fourier Transform (DFT), which efficiently aligns fMRI voxels to fine-grained information of CLIP. We elaborately design a DFT backbone with Spectrum Compression and Frequency Projector modules to learn informative and robust voxel embeddings. Our experiments demonstrate that Lite-Mind achieves an impressive 94.6% fMRI-to-image retrieval accuracy on the NSD dataset for Subject 1, with 98.7% fewer parameters than MindEye. Lite-Mind is also proven to be able to be migrated to smaller fMRI datasets and establishes a new state-of-the-art for zero-shot classification on the GOD dataset.

cs.CV↗

arXiv2024

Learning Variable Impedance Skills from Demonstrations with Passivity Guarantee

Robots are increasingly being deployed not only in workplaces but also in households. Effectively execute of manipulation tasks by robots relies on variable impedance control with contact forces. Furthermore, robots should possess adaptive capabilities to handle the considerable variations exhibited by different robotic tasks in dynamic environments, which can be obtained through human demonstrations. This paper presents a learning-from-demonstration framework that integrates force sensing and motion information to facilitate variable impedance control. The proposed approach involves the estimation of full stiffness matrices from human demonstrations, which are then combined with sensed forces and motion information to create a model using the non-parametric method. This model allows the robot to replicate the demonstrated task while also responding appropriately to new task conditions through the use of the state-dependent stiffness profile. Additionally, a novel tank based variable impedance control approach is proposed to ensure passivity by using the learned stiffness. The proposed approach was evaluated using two virtual variable stiffness systems. The first evaluation demonstrates that the stiffness estimated approach exhibits superior robustness compared to traditional methods when tested on manual datasets, and the second evaluation illustrates that the novel tank based approach is more easily implementable compared to traditional variable impedance control approaches.

cs.RO↗

arXiv2024

Morphing median fin enhances untethered bionic robotic tuna's linear acceleration and turning maneuverability

Median fins of fish-like swimmers play a crucial role in linear acceleration and maneuvering processes. However, few research focused on untethered robotic fish experiments. Imitating the behaviour of real tuna, we developed a free-swimming bionic tuna with a foldable dorsal fin. The erection of dorsal fin, at proper conditions, can reduce head heave by 50%, enhance linear acceleration by 15.7%, increase turning angular velocity by 32.78%, and turning radius decreasing by 33.13%. Conversely, erecting the dorsal fin increases the wetted surface area, resulting in decreased maximum speed and efficiency during steady swimming phase. This finding partially explains why tuna erect their median fins during maneuvers or acceleration and fold them afterward to reduce drag. In addition, we verified that folding the median fins after acceleration does not significantly affect locomotion efficiency. This study supports the application of morphing median fins in undulating underwater robots and helps to further understand the impact of median fins on fish locomotion.

cs.RO↗

arXiv2024

Load Restoration in Islanded Microgrids: Formulation and Solution Strategies

Adverse circumstances such as extreme weather events can cause significant disruptions to normal operation of electric distribution systems (DS), which includes isolating parts of the DS due to damaged transmission equipment. In this paper, we consider the problem of load restoration in a microgrid (MG) that is islanded from the upstream DS. The MG contains sources of distributed generation such as microturbines and renewable energy sources, as well as energy storage systems (ESS). We formulate the load restoration task as a non-convex optimization problem. This problem embodies the physics of the MG by leveraging a branch flow model, while incorporating salient phenomenon in islanded MGs such as the need for internal frequency regulation, and complementarity requirements arising in ESS operations. Since the formulated optimization problem is non-convex, we introduce a convex relaxation which can be solved through model predictive control as a baseline method. However, in order to solve the problem considering its full non-convexity, we leverage a policy-learning method called constrained policy optimization, a tailored version of which is used as our proposed algorithm. The aforementioned approaches, along with an additional deep learning method are compared through extensive simulations.

eess.SY↗

arXiv2024

Discrete Linear Canonical Transform on Graphs: Uncertainty Principle and Sampling

With an increasing influx of classical signal processing methodologies into the field of graph signal processing, approaches grounded in discrete linear canonical transform have found application in graph signals. In this paper, we initially propose the uncertainty principle of the graph linear canonical transform (GLCT), which is based on a class of graph signals maximally concentrated in both vertex and graph spectral domains. Subsequently, leveraging the uncertainty principle, we establish conditions for recovering bandlimited signals of the GLCT from a subset of samples, thereby formulating the sampling theory for the GLCT. We elucidate interesting connections between the uncertainty principle and sampling. Further, by employing sampling set selection and experimental design sampling strategies, we introduce optimal sampling operators in the GLCT domain. Finally, we evaluate the performance of our methods through simulations and numerical experiments across applications.

math.GM↗

arXiv2024

IG Captioner: Information Gain Captioners are Strong Zero-shot Classifiers

Generative training has been demonstrated to be powerful for building visual-language models. However, on zero-shot discriminative benchmarks, there is still a performance gap between models trained with generative and discriminative objectives. In this paper, we aim to narrow this gap by improving the efficacy of generative training on classification tasks, without any finetuning processes or additional modules. Specifically, we focus on narrowing the gap between the generative captioner and the CLIP classifier. We begin by analysing the predictions made by the captioner and classifier and observe that the caption generation inherits the distribution bias from the language model trained with pure text modality, making it less grounded on the visual signal. To tackle this problem, we redesign the scoring objective for the captioner to alleviate the distributional bias and focus on measuring the gain of information brought by the visual inputs. We further design a generative training objective to match the evaluation objective. We name our model trained and evaluated from the novel procedures as Information Gain (IG) captioner. We pretrain the models on the public Laion-5B dataset and perform a series of discriminative evaluations. For the zero-shot classification on ImageNet, IG captioner achieves $> 18\%$ improvements over the standard captioner, achieving comparable performances with the CLIP classifier. IG captioner also demonstrated strong performance on zero-shot image-text retrieval tasks on MSCOCO and Flickr30K. We hope this paper inspires further research towards unifying generative and discriminative training procedures for visual-language models.

cs.CV↗

arXiv2024

MTMamba: Enhancing Multi-Task Dense Scene Understanding by Mamba-Based Decoders

Multi-task dense scene understanding, which learns a model for multiple dense prediction tasks, has a wide range of application scenarios. Modeling long-range dependency and enhancing cross-task interactions are crucial to multi-task dense prediction. In this paper, we propose MTMamba, a novel Mamba-based architecture for multi-task scene understanding. It contains two types of core blocks: self-task Mamba (STM) block and cross-task Mamba (CTM) block. STM handles long-range dependency by leveraging Mamba, while CTM explicitly models task interactions to facilitate information exchange across tasks. Experiments on NYUDv2 and PASCAL-Context datasets demonstrate the superior performance of MTMamba over Transformer-based and CNN-based methods. Notably, on the PASCAL-Context dataset, MTMamba achieves improvements of +2.08, +5.01, and +4.90 over the previous best methods in the tasks of semantic segmentation, human parsing, and object boundary detection, respectively. The code is available at https://github.com/EnVision-Research/MTMamba.

cs.CV↗

arXiv2024

Learning Multiplex Representations on Text-Attributed Graphs with One Language Model Encoder

In real-world scenarios, texts in a graph are often linked by multiple semantic relations (e.g., papers in an academic graph are referenced by other publications, written by the same author, or published in the same venue), where text documents and their relations form a multiplex text-attributed graph. Mainstream text representation learning methods use pretrained language models (PLMs) to generate one embedding for each text unit, expecting that all types of relations between texts can be captured by these single-view embeddings. However, this presumption does not hold particularly in multiplex text-attributed graphs. Along another line of work, multiplex graph neural networks (GNNs) directly initialize node attributes as a feature vector for node representation learning, but they cannot fully capture the semantics of the nodes' associated texts. To bridge these gaps, we propose METAG, a new framework for learning Multiplex rEpresentations on Text-Attributed Graphs. In contrast to existing methods, METAG uses one text encoder to model the shared knowledge across relations and leverages a small number of parameters per relation to derive relation-specific representations. This allows the encoder to effectively capture the multiplex structures in the graph while also preserving parameter efficiency. We conduct experiments on nine downstream tasks in five graphs from both academic and e-commerce domains, where METAG outperforms baselines significantly and consistently. The code is available at https://github.com/PeterGriffinJin/METAG.

cs.CL↗

arXiv2024

Stabilizing Dynamic Systems through Neural Network Learning: A Robust Approach

Point-to-point and periodic motions are ubiquitous in the world of robotics. To master these motions, Autonomous Dynamic System (DS) based algorithms are fundamental in the domain of Learning from Demonstration (LfD). However, these algorithms face the significant challenge of balancing precision in learning with the maintenance of system stability. This paper addresses this challenge by presenting a novel ADS algorithm that leverages neural network technology. The proposed algorithm is designed to distill essential knowledge from demonstration data, ensuring stability during the learning of both point-to-point and periodic motions. For point-to-point motions, a neural Lyapunov function is proposed to align with the provided demonstrations. In the case of periodic motions, the neural Lyapunov function is used with the transversal contraction to ensure that all generated motions converge to a stable limit cycle. The model utilizes a streamlined neural network architecture, adept at achieving dual objectives: optimizing learning accuracy while maintaining global stability. To thoroughly assess the efficacy of the proposed algorithm, rigorous evaluations are conducted using the LASA dataset and a manually designed dataset. These assessments were complemented by empirical validation through robotic experiments, providing robust evidence of the algorithm's performance

cs.RO↗

arXiv2024

AUITestAgent: Automatic Requirements Oriented GUI Function Testing

The Graphical User Interface (GUI) is how users interact with mobile apps. To ensure it functions properly, testing engineers have to make sure it functions as intended, based on test requirements that are typically written in natural language. While widely adopted manual testing and script-based methods are effective, they demand substantial effort due to the vast number of GUI pages and rapid iterations in modern mobile apps. This paper introduces AUITestAgent, the first automatic, natural language-driven GUI testing tool for mobile apps, capable of fully automating the entire process of GUI interaction and function verification. Since test requirements typically contain interaction commands and verification oracles. AUITestAgent can extract GUI interactions from test requirements via dynamically organized agents. Then, AUITestAgent employs a multi-dimensional data extraction strategy to retrieve data relevant to the test requirements from the interaction trace and perform verification. Experiments on customized benchmarks demonstrate that AUITestAgent outperforms existing tools in the quality of generated GUI interactions and achieved the accuracy of verifications of 94%. Moreover, field deployment in Meituan has shown AUITestAgent's practical usability, with it detecting 4 new functional bugs during 10 regression tests in two months.

cs.SE↗

arXiv2024

A First-Order Multi-Gradient Algorithm for Multi-Objective Bi-Level Optimization

In this paper, we study the Multi-Objective Bi-Level Optimization (MOBLO) problem, where the upper-level subproblem is a multi-objective optimization problem and the lower-level subproblem is for scalar optimization. Existing gradient-based MOBLO algorithms need to compute the Hessian matrix, causing the computational inefficient problem. To address this, we propose an efficient first-order multi-gradient method for MOBLO, called FORUM. Specifically, we reformulate MOBLO problems as a constrained multi-objective optimization (MOO) problem via the value-function approach. Then we propose a novel multi-gradient aggregation method to solve the challenging constrained MOO problem. Theoretically, we provide the complexity analysis to show the efficiency of the proposed method and a non-asymptotic convergence result. Empirically, extensive experiments demonstrate the effectiveness and efficiency of the proposed FORUM method in different learning problems. In particular, it achieves state-of-the-art performance on three multi-task learning benchmark datasets. The code is available at https://github.com/Baijiong-Lin/FORUM.

cs.LG↗