arXiv subjects

Qing Gao

Publications and source records attributed to Qing Gao.

At least 19 recordsLinked to original sources

arXiv2026

LightSplat: Real-Time High-Fidelity 3D Gaussian SLAM with Loop Closure

SLAM systems based on 3D Gaussian Splatting (3DGS) have recently demonstrated promising reconstruction accuracy for dense 3D scene representations. However, current 3DGS systems struggle to meet the strict demands of real-world deployments due to severe limitations in operational performance and map adaptability. To this end, we propose LightSplat, a hybrid-representation RGB-D SLAM framework. It synergizes local sparse features for robust and fast tracking with a dual-thread backend that progressively constructs dense Gaussian submaps. Crucially, we enable online loop closure through feature-accelerated 3DGS registration, refining overall map consistency through pose graph optimization. Ultimately, LightSplat achieves the online reconstruction of high-fidelity Gaussian map. Extensive experiments on multiple datasets and real-world robotic platform demonstrate that our method achieves near state-of-the-art reconstruction quality and the capability to accommodate practical camera motions, maintaining an average framerate of 8 FPS. Overall, LightSplat provides an efficient and robust foundation for deploying high-fidelity 3DGS in real-world environments.

cs.RO↗

arXiv2026

DexTele: A Dual-Arm Dexterous Teleoperation System Based on Motion Retargeting and Adaptive Force Control

In dual-arm dexterous teleoperation, cross-platform generalization of motion retargeting and interactivity of grasping are crucial. However, the heterogeneity of robotic architectures and the wide variety of grasping objects pose significant challenges to achieving precise motion retargeting and compliant grasping in dual-arm dexterous teleoperation. To address these challenges, a dual-arm dexterous teleoperation system (DexTele) is proposed based on motion retargeting and adaptive force control. First, a vision-based motion retargeting module is designed to generate preliminary robot motions from human images. In this module, a motion-graph encoder and latent optimization are proposed for precise and convenient cross-platform motion retargeting. Second, an adaptive grasping module is designed to achieve compliant grasping. This module combines a vision-language model (VLM) with model predictive control (MPC), allowing the system to predict the required grasping force for a target object and perform gradient-based online optimization. Finally, extensive experiments demonstrate that the DexTele achieves precise motion retargeting and compliant grasping with generalization across multiple robot platforms.

cs.RO↗

arXiv2026

ObjRetarget: An Object-Aware Motion Retargeting Framework with Anthropomorphic Arm Constraints and Polyhedral Hand Modeling

Learning robot dexterous manipulation from human manipulation videos requires reliably retargeting human intent to executable robot actions while maintaining stable hand-object contact, which remains a key challenge in embodied intelligence. Existing retargeting methods often ignore explicit contact modeling or rely on reinforcement learning, resulting in limited accuracy and generalization. To address this, we propose ObjRetarget, a human-to-robot motion retargeting framework for learning robot dexterous manipulation from human videos, which integrates anthropomorphic arm trajectory constraints with structured hand-object geometric modeling. For arm motion, reference trajectories extracted from human videos are used for initialization, followed by anthropomorphic constraints and redundancy-aware optimization to generate natural and accurate movements. For hand manipulation, ObjRetarget represents multi-finger contacts using polytope clusters and preserves contact structure through geometric invariants to improve stability. Experiments on real robots show that ObjRetarget improves manipulation success rates and contact stability across multiple dexterous tasks, and generalizes well to different demonstrations, object poses, and task settings.

cs.RO↗

arXiv2026

Triple Spectral Fusion for Sensor-based Human Activity Recognition

The field of sensor-based human activity recognition (HAR) mainly uses posture, motion and context data of Inertial Measurement Units (IMUs) to identify daily activities. Despite the advancements in learning-based methods, it is challenging to perform information fusion from the temporal perspective due to the complexities in fusing heterogeneous sensor data and establishing long-term context correlations. This paper proposes a novel triple spectral fusion framework tailored for HAR. First, we develop an adaptive complementary filtering technique for noise suppression and organize each IMU's sensors into posture and motion modality nodes. Given that IMU nodes form a dynamic heterogeneous graph, we then apply adaptive filtering within the graph Fourier domain to merge both homogeneous and heterogeneous node information. Furthermore, an adaptive wavelet frequency selection approach is implemented to suppress context redundancy and shorten the length of features. This approach enhances both timestamp-based graph aggregation and the correlation of long-term contexts. Our framework uses adaptive filtering in the Fourier, graph Fourier, and wavelet domains, enabling effective multi-sensor fusion and context correlation. Extensive experiments on ten benchmark datasets demonstrate the superior performance of our framework. Project page: https://github.com/crocodilegogogo/TSF-TPAMI2026.

cs.AI↗

arXiv2026

MagicAgent: Towards Generalized Agent Planning

The evolution of Large Language Models (LLMs) from passive text processors to autonomous agents has established planning as a core component of modern intelligence. However, achieving generalized planning remains elusive, not only by the scarcity of high-quality interaction data but also by inherent conflicts across heterogeneous planning tasks. These challenges result in models that excel at isolated tasks yet struggle to generalize, while existing multi-task training attempts suffer from gradient interference. In this paper, we present \textbf{MagicAgent}, a series of foundation models specifically designed for generalized agent planning. We introduce a lightweight and scalable synthetic data framework that generates high-quality trajectories across diverse planning tasks, including hierarchical task decomposition, tool-augmented planning, multi-constraint scheduling, procedural logic orchestration, and long-horizon tool execution. To mitigate training conflicts, we propose a two-stage training paradigm comprising supervised fine-tuning followed by multi-objective reinforcement learning over both static datasets and dynamic environments. Empirical results show that MagicAgent-32B and MagicAgent-30B-A3B achieve superior performance across diverse open-source benchmarks (\emph{e.g.}, $75.1\%$ on Worfbench and $86.9\%$ on BFCL-v3), as well as strong results on our in-house MagicEval benchmarks, substantially outperforming existing sub-100B models and surpassing leading ultra-scale models, including GPT-5.2, Kimi-K2 and GLM-4.7.

cs.AI↗

arXiv2026

Adaptive controllable architecture of analog Ising machine

As a quantum-inspired, non-traditional analog solver architecture, the analog Ising machine (AIM) has emerged as a distinctive computational paradigm to address the rapidly growing demand for computational power. However, the mathematical understanding of its principles, as well as the optimization of its solution speed and accuracy, remain unclear. In this work, we for the first time systematically discuss multiple implementations of AIM and establish a unified mathematical formulation. On this basis, by treating the binarization constraint of AIM (such as injection locking) as a Lagrange multiplier in optimization theory and combining it with a Lyapunov analysis from dynamical systems theory, an analytical framework for evaluating solution speed and accuracy is constructed, and further demonstrate that conventional AIMs possess a theoretical performance upper bound. Subsequently, by elevating the binarization constraint to a control variable, we propose the controllable analog Ising machine (CAIM), which integrates control Lyapunov functions and momentum-based optimization algorithms to realize adaptive sampling-feedback control, thereby surpassing the performance limits of conventional AIMs. In a proof-of-concept CAIM demonstration implemented using an FPGA-controlled LC-oscillator Ising machine, CAIM achieves a twofold speedup and a 7\% improvement in accuracy over AIM on a 50-node all-to-all weighted MaxCut problem, validating both the effectiveness and interpretability of the proposed theoretical framework.

quant-ph↗

arXiv2026

DScheLLM: Enabling Dynamic Scheduling through a Fine-Tuned Dual-System Large language Model

Production scheduling is highly susceptible to dynamic disruptions, such as variations in processing times, machine availability, and unexpected task insertions. Conventional approaches typically rely on event-specific models and explicit analytical formulations, which limits their adaptability and generalization across previously unseen disturbances. To overcome these limitations, this paper proposes DScheLLM, a dynamic scheduling approach that leverages fine-tuned large language models within a dual-system (fast-slow) reasoning architecture to address disturbances of different scales. A unified large language model-based framework is constructed to handle dynamic events, where training datasets for both fast and slow reasoning modes are generated using exact schedules obtained from an operations research solver. The Huawei OpenPangu Embedded-7B model is subsequently fine-tuned under the hybrid reasoning paradigms using LoRA. Experimental evaluations on standard job shop scheduling benchmarks demonstrate that the fast-thinking mode can efficiently generate high-quality schedules and the slow-thinking mode can produce solver-compatible and well-formatted decision inputs. To the best of our knowledge, this work represents one of the earliest studies applying large language models to job shop scheduling in dynamic environments, highlighting their considerable potential for intelligent and adaptive scheduling optimization.

cs.AI↗

arXiv2025

$O_k$ null test with multi-task Gaussian processes: cosmic curvature and data compatibility

The $O_k$ null test can not only assess whether the cosmic curvature is zero, therefore if true reducing degeneracies between cosmic curvature and other cosmological parameters, but also provide a model-independent check of compatibility between different data sets. However, traditional implementations often require absolute distance data from Type Ia supernovae (SNe Ia) or baryon acoustic oscillation (BAO) measurements, limiting their applicability because such absolute distance data usually are not accessible. The BAO Alcock Paczynski (AP) parameter $F_{AP}$ is a measurement of a distance ratio, making the Dark Energy Spectroscopic Instrument (DESI) AP measurements particularly well suited for the $O_k$ null test because no absolute distance measurements are required. We propose a novel null test of cosmic curvature tailored to DESI BAO data that combines $F_{AP}$ with ratios such as $D_V'/D_V$ or $D_M'/D_M$. Crucially, this construction eliminates the need for absolute distance measurements. We further develop multi-task Gaussian processes to perform the null test. This approach can also be applied to a joint DESI BAO and SNe Ia dataset, and we find that DESI BAO and SNe Ia data are compatible. Although there is $\sim 2\sigma$ evidence of nonzero curvature at low redshift $z\lesssim 0.5$, this result is not conclusive largely due to the lack of observational data in the corresponding redshift range.

astro-ph.CO↗

arXiv2025

Inflationary Models with Gauss-Bonnet Coupling in Light of ACT Observations

Recent analyses combining Atacama Cosmology Telescope (ACT) data with other cosmological datasets report a higher scalar spectral index $n_s$, creating tension with a wide range of inflationary models. Since a Gauss-Bonnet term with a coupling function $\xi(\phi) = 3\lambda/[4V(\phi)]$ leaves $n_s$ nearly unchanged (up to a field rescaling) while reducing the tensor-to-scalar ratio $r$ by a factor $(1-\lambda)$, so choosing $(1-\lambda)$ sufficiently small effectively removes $r$ as a limiting observable, making it easier for inflationary models to satisfy the latest observational constraints and alleviating this tension. Applying this mechanism to chaotic inflation, E-models, T-models, and hilltop inflation, we find that broad regions of parameter space become consistent with the latest ACT-based CMB constraints. These results demonstrate that Gauss-Bonnet couplings can help bring a broad class of inflationary models into agreement with current CMB measurements.

astro-ph.CO↗

arXiv2025

Observational constraints on inflationary models with non-minimally derivative coupling by ACT

The most recent data release from the Atacama Cosmology Telescope (ACT) reveals a larger value of the scalar spectral tilt $n_s$, ruling out a broad class of inflationary attractors. In this paper, we consider inflationary models including the power law potential, the hilltop model, the polynomial $\alpha$-attractor and exponential $\alpha-$attractor, with non-minimally derivative coupling in the high friction limit, and show how the models can fit ACT data. We also derive constraints on the model parameters using the latest ACT data.

gr-qc↗

arXiv2025

Gravitational wave cosmology

Gravitational waves (GWs) originating from cosmological sources offer direct insights into the physics of the primordial Universe, the fundamental nature of gravity, and the cosmic expansion of the Universe. In this review paper, we present a comprehensive overview of our recent advances in GW cosmology, supported by the national key research and development program of China, focusing on cosmological GW sources and their implications for fundamental physics and cosmology. We first discuss the generation mechanisms and characteristics of stochastic gravitational wave backgrounds generated by physical processes occurred in the early Universe, including those from inflation, phase transitions, and topological defects, and summarize current and possible future constraints from pulsar timing array and space-based detectors. Next, we explore the formation and observational prospects of primordial black holes as GW sources and their potential connection to dark matter. We then analyze how GWs are affected by large-scale structure, cosmological perturbations, and possible modifications of gravity on GW propagation, and how these effects can be used to test fundamental symmetry of gravity. Finally, we discuss the application of GW standard sirens in measuring the Hubble constant, the expansion history, and dark energy parameters, including their combination with electromagnetic observations. These topics together show how GW observations, especially with upcoming space-based detectors, such as LISA, Taiji, and Tianqin, can provide new information about the physics of the early Universe, cosmological evolution, and the nature of gravity.

gr-qc↗

arXiv2025

Approximate reconstruction of inflationary potential with ACT observations

The Atacama Cosmology Telescope (ACT) has recently reported updated measurements of the scalar spectral index $n_s$, revealing a tension with the predictions of many conventional inflationary models. In this work, we adopt a parameterization of the spectral index in the form $n_s = 1 - p/(N + \alpha)$ with $1.338<p<1.746$, to reconstruct an inflationary potential consistent with the latest ACT data. The resulting potential is the Kachru-Kallosh-Linde-Trivedi (KKLT) potential $V(\phi) = V_0/[1 + (M/\phi)^n]$, where the power index is given by $n = 2(p-1)/(2-p)$. The corresponding tensor-to-scalar ratio is approximately $r \approx 16(p-1)/[C (N+\alpha)^p]$ with $C= 2^{2p-1} \left[\sqrt{p-1}/(2-p)\right]^{2p-2} /M^{2p-2}$. Since the reconstruction under the slow-roll approximation is model-independent, the KKLT model can serve as an effective approximation to a broad class of inflationary scenarios that are consistent with the latest ACT measurements of $n_s$, at least on large scales.

astro-ph.CO↗

arXiv2025

Non-minimal coupling in light of ACT

The latest ACT data release disfavors the attractor $n_s=1-2/N$. In inflationary models with nonminimal coupling, such attractors typically arise in the strong coupling limit. To align with observational constraints, we focus on nonminimal coupling models with small coupling constants. For the model with the coupling function $\Omega(\phi) = 1 + \xi f(\phi)$ and the potential $V(\phi) = \lambda^2 f^2(\phi)$, we find that observational data constrain the parameters as $0.1 \lesssim \xi \lesssim 35$ and $0 \lesssim k \lesssim 1.5$ for $f(\phi) = \phi^k$ at the $1\sigma$ confidence level. With the help of the nonmiminal coupling $\Omega(\phi) = 1 + \xi \phi^2$, the hilltop inflation and power-law inflation models with power indices $2/3$ and $1/3$ can be consistent with observational data within the $1\sigma$ range. We also give the viable parameter regions for $\xi$ for these three models.

astro-ph.CO↗

arXiv2025

Null tests with Gaussian Process

We investigate the null tests of spatial flatness and the flat $\Lambda$CDM model using the Baryon Acoustic Oscillation (BAO) data measured by the Dark Energy Spectroscopic Instrument (DESI), the cosmic chronometers (CCH) $H(z)$ data, and the Union3 and Pantheon Plus type Ia supernovae (SNe Ia) datasets. We propose a novel non-parametric reconstruction of $F_{AP}$, $D_M/r_d$ and $D'_M/r_d$ from the DESI BAO data to perform the $Ok$ diagnostic, and we also conduct the $Ok$ diagnostic using the combination of CCH and SNe Ia data. The novel method avoids the issue of the dependence on cosmological parameters such as the value of the Hubble constant. There is no evidence of deviation from the flat $\Lambda$CDM model, nor is there any indication of dynamical dark energy found in the observational data. Since we employ a non-parametric reconstruction method, all the conclusions drawn in this paper remain robust and agnostic to any cosmological model and gravitational theory.

astro-ph.CO↗

arXiv2025

The Nature of Classical Be Star Outbursts: A Multi-Epoch Study of the Be system EPIC 202060631

Through cross-matching the ASASSN photometric light curves and LAMOST spectroscopic observations, we serendipitously captured a rare major outburst event from the Be star EPIC 202060631, lasting over 1000 days. Fortuitously, the LAMOST ow-resolution spectra densely covered the flux rising stage with 20 epochs, while an additional 7 low-resolution and 11 medium esolution spectra monitored the subsequent decay phase. Moreover, this target was observed by Kepler telescope in its K2 mission when before the outburst and by TESS telescope while after returning to quiescence. Analyses of these datasets reveal pulsation behavior and mode amplitudes, indications of radial and tangential motions in the photosphere, and clear evidence of mass ejection and circumstellar disk formation. We modeled the central star using the BRUCE04 code and analyzed the disk structure using HDUST. Our results suggest episodic mass injection events from the central star triggered the disk buildup and carried imprints of the changing stellar pulsations. This study offers unique insights into the connections between photospheric activities, disk evolution, and stellar rotation during Be star outbursts.

astro-ph.SR↗

arXiv2024

A Powered Prosthetic Hand with Vision System for Enhancing the Anthropopathic Grasp

The anthropomorphism of grasping process significantly benefits the experience and grasping efficiency of prosthetic hand wearers. Currently, prosthetic hands controlled by signals such as brain-computer interfaces (BCI) and electromyography (EMG) face difficulties in precisely recognizing the amputees' grasping gestures and executing anthropomorphic grasp processes. Although prosthetic hands equipped with vision systems enables the objects' feature recognition, they lack perception of human grasping intention. Therefore, this paper explores the estimation of grasping gestures solely through visual data to accomplish anthropopathic grasping control and the determination of grasping intention within a multi-object environment. To address this, we propose the Spatial Geometry-based Gesture Mapping (SG-GM) method, which constructs gesture functions based on the geometric features of the human hand grasping processes. It's subsequently implemented on the prosthetic hand. Furthermore, we propose the Motion Trajectory Regression-based Grasping Intent Estimation (MTR-GIE) algorithm. This algorithm predicts pre-grasping object utilizing regression prediction and prior spatial segmentation estimation derived from the prosthetic hand's position and trajectory. The experiments were conducted to grasp 8 common daily objects including cup, fork, etc. The experimental results presented a similarity coefficient $R^{2}$ of grasping process of 0.911, a Root Mean Squared Error ($RMSE$) of 2.47\degree, a success rate of grasping of 95.43$\%$, and an average duration of grasping process of 3.07$\pm$0.41 s. Furthermore, grasping experiments in a multi-object environment were conducted. The average accuracy of intent estimation reached 94.35$\%$. Our methodologies offer a groundbreaking approach to enhance the prosthetic hand's functionality and provides valuable insights for future research.

cs.RO↗

arXiv2024

Quantum Discrete Adiabatic Linear Solver based on Block Encoding and Eigenvalue Separator

Linear system solvers are widely used in scientific computing, with the primary goal of solving linear system problems. Classical iterative algorithms typically rely on the conjugate gradient method. The rise of quantum computing has spurred interest in quantum linear system problems (QLSP), particularly following the introduction of the HHL algorithm by Harrow et al. in 2009, which demonstrated the potential for exponential speedup compared to classical algorithms. However, the performance of the HHL algorithm is constrained by its dependence on the square of the condition number. To address this limitation, alternative approaches based on adiabatic quantum computing (AQC) have been proposed, which exhibit complexity scaling linearly with the condition number. AQC solves QLSP by smoothly varying the parameters of the Hamiltonian. However, this method suffers from high Hamiltonian simulation complexity. In response, this work designs new Hamiltonians and proposes a quantum discrete adiabatic linear solver based on block encoding and eigenvalue separation techniques (BEES-QDALS). This approach bypasses Hamiltonian simulation through a first-order approximation and leverages block encoding to achieve equivalent non-unitary operations on qubits. By comparing the fidelity of the original algorithm and BEES-QDALS when solving QLSP with a fixed number of steps, and the number of steps required to reach a target fidelity, it is found that BEES-QDALS significantly outperforms the original algorithm. Specifically, BEES-QDALS achieves higher fidelity with the same number of steps and requires fewer steps to reach the same target fidelity.

quant-ph↗

arXiv2024

On the evidence of dynamical dark energy

To elucidate the robustness of the baryon acoustic oscillation (BAO) data measured by the Dark Energy Spectroscopic Instrument (DESI) in capturing the dynamical behavior of dark energy, we assess the model dependence of the evidence for dynamical dark energy inferred from the DESI BAO data. While the DESI BAO data slightly tightens the constraints on model parameters and increases the tension between the Chevallier-Polarski-Linder (CPL) model and the $\Lambda$CDM model, we find that the influence of DESI BAO data on the constraint of $w_0$ is small in the SSLCPL model. In comparison to the CPL model, the tension with the $\Lambda$CDM model is reduced for the SSLCPL model, suggesting that the evidence for dynamical dark energy from DESI BAO data is dependent on cosmological models. The inclusion of spatial curvature has little impact on the results in the SSLCPL model.

astro-ph.CO↗