arXiv subjects

Bolin Li

Publications and source records attributed to Bolin Li.

15 recordsLinked to original sources

arXiv2025

Generative Multi-Focus Image Fusion

Multi-focus image fusion aims to generate an all-in-focus image from a sequence of partially focused input images. Existing fusion algorithms generally assume that, for every spatial location in the scene, there is at least one input image in which that location is in focus. Furthermore, current fusion models often suffer from edge artifacts caused by uncertain focus estimation or hard-selection operations in complex real-world scenarios. To address these limitations, we propose a generative multi-focus image fusion framework, termed GMFF, which operates in two sequential stages. In the first stage, deterministic fusion is implemented using StackMFF V4, the latest version of the StackMFF series, and integrates the available focal plane information to produce an initial fused image. The second stage, generative restoration, is realized through IFControlNet, which leverages the generative capabilities of latent diffusion models to reconstruct content from missing focal planes, restore fine details, and eliminate edge artifacts. Each stage is independently developed and functions seamlessly in a cascaded manner. Extensive experiments demonstrate that GMFF achieves state-of-the-art fusion performance and exhibits significant potential for practical applications, particularly in scenarios involving complex multi-focal content. The implementation is publicly available at https://github.com/Xinzhe99/StackMFF-Series.

cs.CV↗

arXiv2025

Study of Neutron Star Properties under the Two-Flavor Quark NJL Model

The Equation of State (EOS) of matter within neutron stars is a central topic in nuclear physics and astrophysics.This study investigates hadron-quark hybrid stars by integrating the density-dependent DDME2 relativistic mean-field model for hadronic matter with a two-flavor Nambu-Jona-Lasinio (NJL) model for quark matter.A quintic polynomial interpolation is employed to construct a smooth ($C^2$ continuity) and thermodynamically consistent crossover between the phases.We systematically explore the parameter space to reconcile the tension between the high stiffness required by massive pulsars and the softness demanded by tidal deformability and radius constraints.Our analysis demonstrates that to simultaneously satisfy the mass measurement of PSR J0740+6620 and the compact radius constraints from NICER (e.g., PSR J0437-4715), the hadron-quark crossover must initiate in the vicinity of nuclear saturation density.This result suggests that the early percolation of quark degrees of freedom is a necessary feature to accommodate current multi-messenger observations.

astro-ph.HE↗

arXiv2025

A Three-Level Whole-Body Disturbance Rejection Control Framework for Dynamic Motions in Legged Robots

This paper presents a control framework designed to enhance the stability and robustness of legged robots in the presence of uncertainties, including model uncertainties, external disturbances, and faults. The framework enables the full-state feedback estimator to estimate and compensate for uncertainties in the whole-body dynamics of the legged robots. First, we propose a novel moving horizon extended state observer (MH-ESO) to estimate uncertainties and mitigate noise in legged systems, which can be integrated into the framework for disturbance compensation. Second, we introduce a three-level whole-body disturbance rejection control framework (T-WB-DRC). Unlike the previous two-level approach, this three-level framework considers both the plan based on whole-body dynamics without uncertainties and the plan based on dynamics with uncertainties, significantly improving payload transportation, external disturbance rejection, and fault tolerance. Third, simulations of both humanoid and quadruped robots in the Gazebo simulator demonstrate the effectiveness and versatility of T-WB-DRC. Finally, extensive experimental trials on a quadruped robot validate the robustness and stability of the system when using T-WB-DRC under various disturbance conditions.

cs.RO↗

arXiv2025

A Gait Driven Reinforcement Learning Framework for Humanoid Robots

This paper presents a real-time gait driven training framework for humanoid robots. First, we introduce a novel gait planner that incorporates dynamics to design the desired joint trajectory. In the gait design process, the 3D robot model is decoupled into two 2D models, which are then approximated as hybrid inverted pendulums (H-LIP) for trajectory planning. The gait planner operates in parallel in real time within the robot's learning environment. Second, based on this gait planner, we design three effective reward functions within a reinforcement learning framework, forming a reward composition to achieve periodic bipedal gait. This reward composition reduces the robot's learning time and enhances locomotion performance. Finally, a gait design example, along with simulation and experimental comparisons, is presented to demonstrate the effectiveness of the proposed method.

cs.RO↗

arXiv2025

Disturbance Estimation of Legged Robots: Predefined Convergence via Dynamic Gains

In this study, we address the challenge of disturbance estimation in legged robots by introducing a novel continuous-time online feedback-based disturbance observer that leverages measurable variables. The distinct feature of our observer is the integration of dynamic gains and comparison functions, which guarantees predefined convergence of the disturbance estimation error, including ultimately uniformly bounded, asymptotic, and exponential convergence, among various types. The properties of dynamic gains and the sufficient conditions for comparison functions are detailed to guide engineers in designing desired convergence behaviors. Notably, the observer functions effectively without the need for upper bound information of the disturbance or its derivative, enhancing its engineering applicability. An experimental example corroborates the theoretical advancements achieved.

cs.RO↗

arXiv2025

Streaming Video Question-Answering with In-context Video KV-Cache Retrieval

We propose ReKV, a novel training-free approach that enables efficient streaming video question-answering (StreamingVQA), by seamlessly integrating with existing Video Large Language Models (Video-LLMs). Traditional VideoQA systems struggle with long videos, as they must process entire videos before responding to queries, and repeat this process for each new question. In contrast, our approach analyzes long videos in a streaming manner, allowing for prompt responses as soon as user queries are received. Building on a common Video-LLM, we first incorporate a sliding-window attention mechanism, ensuring that input frames attend to a limited number of preceding frames, thereby reducing computational overhead. To prevent information loss, we store processed video key-value caches (KV-Caches) in RAM and disk, reloading them into GPU memory as needed. Additionally, we introduce a retrieval method that leverages an external retriever or the parameters within Video-LLMs to retrieve only query-relevant KV-Caches, ensuring both efficiency and accuracy in question answering. ReKV enables the separation of video encoding and question-answering across different processes and GPUs, significantly enhancing the efficiency of StreamingVQA. Through comprehensive experimentation, we validate the efficacy and practicality of our approach, which significantly boosts efficiency and enhances applicability over existing VideoQA models.

cs.CV↗

arXiv2025

A Whole-Body Disturbance Rejection Control Framework for Dynamic Motions in Legged Robots

This letter presents a control framework for legged robots that enables self-perception and resistance to external disturbances and model uncertainties. First, a novel disturbance estimator is proposed, integrating adaptive control and extended state observers (ESO) to estimate external disturbances and model uncertainties. This estimator is embedded within the whole-body control framework to compensate for disturbances in the legged system. Second, a comprehensive whole-body disturbance rejection control framework (WB-DRC) is introduced, accounting for the robot's full-body dynamics. Compared to previous whole-body control frameworks, WB-DRC effectively handles external disturbances and model uncertainties, with the potential to adapt to complex terrain. Third, simulations of both biped and quadruped robots are conducted in the Gazebo simulator to demonstrate the effectiveness and versatility of WB-DRC. Finally, extensive experimental trials on the quadruped robot validate the robustness and stability of the robot system using WB-DRC under various disturbance conditions.

cs.RO↗

arXiv2024

Boosting Private Domain Understanding of Efficient MLLMs: A Tuning-free, Adaptive, Universal Prompt Optimization Framework

Efficient multimodal large language models (EMLLMs), in contrast to multimodal large language models (MLLMs), reduce model size and computational costs and are often deployed on resource-constrained devices. However, due to data privacy concerns, existing open-source EMLLMs rarely have access to private domain-specific data during the pre-training process, making them difficult to directly apply in device-specific domains, such as certain business scenarios. To address this weakness, this paper focuses on the efficient adaptation of EMLLMs to private domains, specifically in two areas: 1) how to reduce data requirements, and 2) how to avoid parameter fine-tuning. Specifically, we propose a tun\textbf{\underline{I}}ng-free, a\textbf{\underline{D}}aptiv\textbf{\underline{E}}, univers\textbf{\underline{AL}} \textbf{\underline{Prompt}} Optimization Framework, abbreviated as \textit{\textbf{\ourmethod{}}} which consists of two stages: 1) Predefined Prompt, based on the reinforcement searching strategy, generate a prompt optimization strategy tree to acquire optimization priors; 2) Prompt Reflection initializes the prompt based on optimization priors, followed by self-reflection to further search and refine the prompt. By doing so, \ourmethod{} elegantly generates the ``ideal prompts'' for processing private domain-specific data. Note that our method requires no parameter fine-tuning and only a small amount of data to quickly adapt to the data distribution of private data. Extensive experiments across multiple tasks demonstrate that our proposed \ourmethod{} significantly improves both efficiency and performance compared to baselines.

cs.AI↗

arXiv2024

LLaVA-MoD: Making LLaVA Tiny via MoE Knowledge Distillation

We introduce LLaVA-MoD, a novel framework designed to enable the efficient training of small-scale Multimodal Language Models (s-MLLM) by distilling knowledge from large-scale MLLM (l-MLLM). Our approach tackles two fundamental challenges in MLLM distillation. First, we optimize the network structure of s-MLLM by integrating a sparse Mixture of Experts (MoE) architecture into the language model, striking a balance between computational efficiency and model expressiveness. Second, we propose a progressive knowledge transfer strategy to ensure comprehensive knowledge migration. This strategy begins with mimic distillation, where we minimize the Kullback-Leibler (KL) divergence between output distributions to enable the student model to emulate the teacher network's understanding. Following this, we introduce preference distillation via Direct Preference Optimization (DPO), where the key lies in treating l-MLLM as the reference model. During this phase, the s-MLLM's ability to discriminate between superior and inferior examples is significantly enhanced beyond l-MLLM, leading to a better student that surpasses its teacher, particularly in hallucination benchmarks. Extensive experiments demonstrate that LLaVA-MoD outperforms existing models across various multimodal benchmarks while maintaining a minimal number of activated parameters and low computational costs. Remarkably, LLaVA-MoD, with only 2B activated parameters, surpasses Qwen-VL-Chat-7B by an average of 8.8% across benchmarks, using merely 0.3% of the training data and 23% trainable parameters. These results underscore LLaVA-MoD's ability to effectively distill comprehensive knowledge from its teacher model, paving the way for the development of more efficient MLLMs. The code will be available on: https://github.com/shufangxun/LLaVA-MoD.

cs.CV↗

arXiv2024

Hybrid Dynamics Modeling and Trajectory Planning for a Cable-Trailer System with a Quadruped Robot

Inspired by sled-pulling dogs in transportation, we present a cable-trailer integrated with a quadruped robot system. The motion planning of this system faces challenges due to the interactions between the cable's state transitions, the trailer's nonholonomic constraints, and the system's underactuation. To address these challenges, we first develop a hybrid dynamics model that captures the cable's taut and slack states. A search algorithm is then introduced to compute a suboptimal trajectory while incorporating mode transitions. Additionally, we propose a novel collision avoidance constraint based on geometric polygons to formulate the trajectory optimization problem for the hybrid system. The proposed method is implemented on a Unitree A1 quadruped robot with a customized cable-trailer and validated through experiments. The real system demonstrates both agile and safe motion with cable mode transitions.

cs.RO↗

arXiv2022

An Improved Active Disturbance Rejection Control for Bode's Ideal Transfer Function

This paper presents an active disturbance rejection control (ADRC) scheme with an improved fractional-order extended state observer (IFO-ESO).Based on the new ADRC scheme, the open-loop transfer function of a high-order system can be approximately rendered to a so-called Weighed Bode's ideal transfer function, whose closed-loop performance is less prone to the controller parameter variations. The design of the IFO-ESO helps reduce the number of system states to be estimated and improves the performance of closed-loop system over %the previously proposed the fractional-order active disturbance rejection control (FO-ADRC) in the literature. Compared with the integer-order active disturbance rejection controller (IO-ADRC) and FO-ADRC, the auxiliary tracking controller of IFO-ADRC has a simpler form. Frequency-domain analysis shows that IFO-ESO has better estimation performance than fractional-order ESO, and time-domain simulation demonstrates that the proposed ADRC has better transient performance and is more robust against the parameter variations than FO-ADRC and IO-ADRC. The proposed ADRC is applied to permanent magnet synchronous motor (PMSM) servo control system and demonstrates its capability in a real-world application.

eess.SY↗

arXiv2021

An Improved Fractional-Order Active Disturbance Rejection Control: Performance Analysis and Experiment Verification

This paper presents an improved active disturbance rejection control scheme (IFO-ADRC) with an improved fractional-order extended state observer (IFO-ESO). The structural information of the system is utilized in IFO-ESO rather than buried as in the typical fractional-order extended state observer (FO-ESO) and help significantly improve the performance of IFO-ESO and closed-loop system. Compared with the integer-order active disturbance rejection controller (IO-ADRC), the auxiliary tracking controller of IFO-ADRC has a simpler form and fewer parameters need to be tuned. Frequency-domain analysis shows that IFO-ESO has better performance over the larger frequency band than FO-ESO, and time-domain simulation shows that IFO-ADRC has better transient performance and is more robust against the parameter variations than traditional fractional-order active disturbance rejection controller (FO-ADRC) and IO-ADRC. The IFO-ADRC is applied to permanent magnet synchronous motor (PMSM) servo control system and demonstrates its capability in the real-world application.

eess.SY↗

arXiv2021

A new active disturbance controller based on an improved fraction-order extended state observer

This paper proposes a new fraction-order active disturbance rejection controller based on an improved fraction-order extended state observer (IFESO) for a class of fraction-order systems. Applying the IFESO, the fraction-order system can be approximated as an integer-order integrator (1/s). The parameters that ensure the stability of the IFESO and closed-loop system are provided. The frequency-domain analysis shows that the IFADRC has a stronger disturbance estimation performance for the fraction-order system than an integer-order active disturbance rejection controller (IADRC). The simulation results demonstrates that the proposed IFADRC behaves better than the IADRC and the FADRC.

eess.SY↗

arXiv2021

CRT-Net: A Generalized and Scalable Framework for the Computer-Aided Diagnosis of Electrocardiogram Signals

Electrocardiogram (ECG) signals play critical roles in the clinical screening and diagnosis of many types of cardiovascular diseases. Despite deep neural networks that have been greatly facilitated computer-aided diagnosis (CAD) in many clinical tasks, the variability and complexity of ECG in the clinic still pose significant challenges in both diagnostic performance and clinical applications. In this paper, we develop a robust and scalable framework for the clinical recognition of ECG. Considering the fact that hospitals generally record ECG signals in the form of graphic waves of 2-D images, we first extract the graphic waves of 12-lead images into numerical 1-D ECG signals by a proposed bi-directional connectivity method. Subsequently, a novel deep neural network, namely CRT-Net, is designed for the fine-grained and comprehensive representation and recognition of 1-D ECG signals. The CRT-Net can well explore waveform features, morphological characteristics and time domain features of ECG by embedding convolution neural network(CNN), recurrent neural network(RNN), and transformer module in a scalable deep model, which is especially suitable in clinical scenarios with different lengths of ECG signals captured from different devices. The proposed framework is first evaluated on two widely investigated public repositories, demonstrating the superior performance of ECG recognition in comparison with state-of-the-art. Moreover, we validate the effectiveness of our proposed bi-directional connectivity and CRT-Net on clinical ECG images collected from the local hospital, including 258 patients with chronic kidney disease (CKD), 351 patients with Type-2 Diabetes (T2DM), and around 300 patients in the control group. In the experiments, our methods can achieve excellent performance in the recognition of these two types of disease.

cs.LG↗

arXiv2014

Tunable Thermal Conduction in Graphane Nanoribbons

Graphane and graphene are both two-dimensional materials but of different bonding configurations, which can result in distinct thermal conduction properties. We simulate thermal conduction in graphane nanoribbons (GANRs) using the nonequilibrium Green's function method. It is found that GANRs have lower ballistic thermal conductance and stronger thermal conductance anisotropy than the graphene counterparts. Furthermore, hydrogen vacancies of GANRs considerably suppress thermal conduction, accompanied by enhanced thermal conductance anisotropy. The tunable thermal conduction, realized by controlling the width, edge shape and hydrogen vacancy concentration of GANRs, could be useful for thermal management and thermoelectric applications.

cond-mat.mes-hall↗