arXiv subjects

Qifeng Zhang

Publications and source records attributed to Qifeng Zhang.

16 recordsLinked to original sources

arXiv2026

Semantically Compatible Knowledge Distillation for Cross-Domain Object Detection with Vision Foundation Models

Vision foundation models (VFMs) offer strong generalization capabilities for domain-adaptive object detection (DAOD). However, existing VFM-based methods overlook the spatial-scale discrepancy between teacher and student feature maps, resulting in semantic incompatibility that weakens both feature alignment and pseudo-label learning. Moreover, domain shift can cause source-trained VFM teachers to miss target-domain objects, limiting the quality of their pseudo-labels. To address these issues, we propose the Semantic Localization-Enhanced Teacher (SLE-T), a semantically compatible knowledge-distillation framework built around a lightweight SLE Adapter for DINOv2. SLE Adapter injects pretrained local-texture priors into DINOv2 to improve cross-domain recognition and reformulates its features into dense representations that are spatially and semantically compatible with the student detector. SLE-T transfers the resulting teacher knowledge through either pseudo-label learning or feature alignment. We instantiate SLE-T with DINOv2-B and DINOv2-L (the ViT-B and ViT-L variants) and compare them with the larger DINOv2-G teacher. Extensive experiments on three DAOD benchmarks demonstrate that our method achieves state-of-the-art performance, and ablation studies confirm the importance of teacher-student semantic compatibility. Notably, SLE-T with DINOv2-B produces competitive or superior pseudo-labels using approximately one-quarter of the training time of DINOv2-G and substantially less GPU memory, demonstrating efficient VFM knowledge transfer under limited computational resources.

cs.CV↗

arXiv2026

GST-Bench: Can VLMs Develop Global Spatial Awareness from Video?

Spatial intelligence is fundamental to embodied agents, yet existing benchmarks focus on local spatial perception from single or few viewpoints, overlooking global spatial awareness over continuous, long-horizon visual streams. To address this limitation, we introduce the Global-Spatial-Temporal Benchmark (GST-Bench), a VQA benchmark for global spatial intelligence in video understanding, comprising human-verified questions derived from 6,790 minutes of synthetically generated video. It requires models to perform accurate spatial inference from novel viewpoints unseen in the input video and to map egocentric observations onto global top-down images. A comprehensive evaluation of 22 state-of-the-art VLMs exposes a striking gap between models and humans: the strongest zero-shot model attains only 42.68, far below the human score of 79.08. To probe the cause of this gap, we construct GST-Bench-Local and find that models, despite strong local spatial understanding under the same task formulation, still fail to consolidate long-horizon observations into a globally consistent scene representation. We further provide GST-Train, a dataset for global spatial reasoning, as a complementary resource to facilitate future research on this challenge.

cs.CV↗

arXiv2025

Robix: A Unified Model for Robot Interaction, Reasoning and Planning

We introduce Robix, a unified model that integrates robot reasoning, task planning, and natural language interaction within a single vision-language architecture. Acting as the high-level cognitive layer in a hierarchical robot system, Robix dynamically generates atomic commands for the low-level controller and verbal responses for human interaction, enabling robots to follow complex instructions, plan long-horizon tasks, and interact naturally with human within an end-to-end framework. Robix further introduces novel capabilities such as proactive dialogue, real-time interruption handling, and context-aware commonsense reasoning during task execution. At its core, Robix leverages chain-of-thought reasoning and adopts a three-stage training strategy: (1) continued pretraining to enhance foundational embodied reasoning abilities including 3D spatial understanding, visual grounding, and task-centric reasoning; (2) supervised finetuning to model human-robot interaction and task planning as a unified reasoning-action sequence; and (3) reinforcement learning to improve reasoning-action consistency and long-horizon task coherence. Extensive experiments demonstrate that Robix outperforms both open-source and commercial baselines (e.g., GPT-4o and Gemini 2.5 Pro) in interactive task execution, demonstrating strong generalization across diverse instruction types (e.g., open-ended, multi-stage, constrained, invalid, and interrupted) and various user-involved tasks such as table bussing, grocery shopping, and dietary filtering.

cs.AI↗

arXiv2025

Enhancing Small-Scale Dataset Expansion with Triplet-Connection-based Sample Re-Weighting

The performance of computer vision models in certain real-world applications, such as medical diagnosis, is often limited by the scarcity of available images. Expanding datasets using pre-trained generative models is an effective solution. However, due to the uncontrollable generation process and the ambiguity of natural language, noisy images may be generated. Re-weighting is an effective way to address this issue by assigning low weights to such noisy images. We first theoretically analyze three types of supervision for the generated images. Based on the theoretical analysis, we develop TriReWeight, a triplet-connection-based sample re-weighting method to enhance generative data augmentation. Theoretically, TriReWeight can be integrated with any generative data augmentation methods and never downgrade their performance. Moreover, its generalization approaches the optimal in the order $O(\sqrt{d\ln (n)/n})$. Our experiments validate the correctness of the theoretical analysis and demonstrate that our method outperforms the existing SOTA methods by $7.9\%$ on average over six natural image datasets and by $3.4\%$ on average over three medical datasets. We also experimentally validate that our method can enhance the performance of different generative data augmentation methods.

cs.CV↗

arXiv2024

FedMPQ: Secure and Communication-Efficient Federated Learning with Multi-codebook Product Quantization

In federated learning, particularly in cross-device scenarios, secure aggregation has recently gained popularity as it effectively defends against inference attacks by malicious aggregators. However, secure aggregation often requires additional communication overhead and can impede the convergence rate of the global model, which is particularly challenging in wireless network environments with extremely limited bandwidth. Therefore, achieving efficient communication compression under the premise of secure aggregation presents a highly challenging and valuable problem. In this work, we propose a novel uplink communication compression method for federated learning, named FedMPQ, which is based on multi shared codebook product quantization.Specifically, we utilize updates from the previous round to generate sufficiently robust codebooks. Secure aggregation is then achieved through trusted execution environments (TEE) or a trusted third party (TTP).In contrast to previous works, our approach exhibits greater robustness in scenarios where data is not independently and identically distributed (non-IID) and there is a lack of sufficient public data. The experiments conducted on the LEAF dataset demonstrate that our proposed method achieves 99% of the baseline's final accuracy, while reducing uplink communications by 90-95%

cs.CR↗

arXiv2023

Error estimates of invariant-preserving difference schemes for the rotation-two-component Camassa--Holm system with small energy

A rotation-two-component Camassa-Holm (R2CH) system was proposed recently to describe the motion of shallow water waves under the influence of gravity. This is a highly nonlinear and strongly coupled system of partial differential equations. A crucial issue in designing numerical schemes is to preserve invariants as many as possible at the discrete level. In this paper, we present a provable implicit nonlinear difference scheme which preserves at least three discrete conservation invariants: energy, mass, and momentum, and prove the existence of the difference solution via the Browder theorem. The error analysis is based on novel and refined estimates of the bilinear operator in the difference scheme. By skillfully using the energy method, we prove that the difference scheme not only converges unconditionally when the rotational parameter diminishes, but also converges without any step-ratio restriction for the small energy case when the rotational parameter is nonzero. The convergence orders in both settings (zero or nonzero rotation parameter) are $O(τ^2 + h^2)$ for the velocity in the $L^\infty$-norm and the surface elevation in the $L^2$-norm, where $τ$ denotes the temporal stepsize and $h$ the spatial stepsize, respectively. The theoretical predictions are confirmed by a properly designed two-level iteration scheme. Comparing with existing numerical methods in the literature, the proposed method demonstrates its effectiveness for long-time simulation over larger domains and superior resolution for both smooth and non-smooth initial values.

math.NA↗

arXiv2023

Fully Conservative Difference Schemes for the Rotation-Two-Component Camassa-Holm System with Smooth/Nonsmooth Initial Data

The rotation-two-component Camassa--Holm system, which possesses strongly nonlinear coupled terms and high-order differential terms, tends to have continuous nonsmooth solitary wave solutions, such as peakons, stumpons, composite waves and even chaotic waves. In this paper an accurate semi-discrete conservative difference scheme for the system is derived by taking advantage of its Hamiltonian invariants. We show that the semi-discrete numerical scheme preserves at least three discrete conservative laws: mass, momentum and energy. Furthermore, a fully discrete finite difference scheme is proposed without destroying anyone of the conservative laws. Combining a nonlinear iteration process and an efficient threshold strategy, the accuracy of the numerical scheme can be guaranteed. Meanwhile, the difference scheme can capture the formation and propagation of solitary wave solutions with satisfying long time behavior under the smooth/nonsmooth initial data. The numerical results reveal a new type of asymmetric wave breaking phenomenon under the nonzero rotational parameter.

math.NA↗

arXiv2023

The energy method for high-order invariants in shallow water wave equations

Third order dispersive evolution equations are widely adopted to model one-dimensional long waves and have extensive applications in fluid mechanics, plasma physics and nonlinear optics. Among them are the KdV equation, the Camassa--Holm equation and the Degasperis--Procesi equation. They share many common features such as complete integrability, Lax pairs and bi-Hamiltonian structure. In this paper we revisit high-order invariants for these three types of shallow water wave equations by the energy method in combination of a skew-adjoint operator $(1-\partial_{xx})^{-1}$. Several applications to seek high-order invariants of the Benjamin-Bona-Mahony equation, the regularized long wave equation and the Rosenau equation are also presented.

math.NA↗

arXiv2022

Optimal convergence rate of the explicit Euler method for convection-diffusion equations II: high dimensional cases

This is the second part of study on the optimal convergence rate of the explicit Euler discretization in time for the convection-diffusion equations [Appl. Math. Lett. \textbf{131} (2022) 108048] which focuses on high-dimensional linear/nonlinear cases under Dirichlet or Neumann boundary conditions. Several new corrected difference schemes are proposed based on the explicit Euler discretization in temporal derivative and central difference discretization in spatial derivatives. The priori estimate of the corrected scheme with application to constant convection coefficients is provided at length by the maximum principle and the optimal convergence rate four is proved when the step ratios along each direction equal to $1/6$. The corrected difference schemes have essentially improved {\rm \textbf{CFL}} condition and the numerical accuracy comparing with the classical difference schemes. Numerical examples involving two-/three-dimensional linear/nonlinear problems under Dirichlet/Neumann boundary conditions such as the Fisher equation, the Chafee-Infante equation, the Burgers' equation and classification to name a few substantiate the good properties claimed for the corrected difference scheme.

math.NA↗

arXiv2022

A Transferable Legged Mobile Manipulation Framework Based on Disturbance Predictive Control

Due to their ability to adapt to different terrains, quadruped robots have drawn much attention in the research field of robot learning. Legged mobile manipulation, where a quadruped robot is equipped with a robotic arm, can greatly enhance the performance of the robot in diverse manipulation tasks. Several prior works have investigated legged mobile manipulation from the viewpoint of control theory. However, modeling a unified structure for various robotic arms and quadruped robots is a challenging task. In this paper, we propose a unified framework disturbance predictive control where a reinforcement learning scheme with a latent dynamic adapter is embedded into our proposed low-level controller. Our method can adapt well to various types of robotic arms with a few random motion samples and the experimental results demonstrate the effectiveness of our method.

cs.RO↗

arXiv2022

Imitation and Adaptation Based on Consistency: A Quadruped Robot Imitates Animals from Videos Using Deep Reinforcement Learning

The essence of quadrupeds' movements is the movement of the center of gravity, which has a pattern in the action of quadrupeds. However, the gait motion planning of the quadruped robot is time-consuming. Animals in nature can provide a large amount of gait information for robots to learn and imitate. Common methods learn animal posture with a motion capture system or numerous motion data points. In this paper, we propose a video imitation adaptation network (VIAN) that can imitate the action of animals and adapt it to the robot from a few seconds of video. The deep learning model extracts key points during animal motion from videos. The VIAN eliminates noise and extracts key information of motion with a motion adaptor, and then applies the extracted movements function as the motion pattern into deep reinforcement learning (DRL). To ensure similarity between the learning result and the animal motion in the video, we introduce rewards that are based on the consistency of the motion. DRL explores and learns to maintain balance from movement patterns from videos, imitates the action of animals, and eventually, allows the model to learn the gait or skills from short motion videos of different animals and to transfer the motion pattern to the real robot.

cs.RO↗

arXiv2021

A conservative difference scheme with optimal pointwise error estimates for two-dimensional space fractional nonlinear Schrödinger equations

In this paper, a linearized semi-implicit finite difference scheme is proposed for solving the two-dimensional (2D) space fractional nonlinear Schrödinger equation (SFNSE).The scheme has the property of mass and energy conservation on the discrete level, with an unconditional stability and a second order accuracy for both time and spatial variables. The main contribution of this paper is an optimal pointwise error estimate for the 2D SFNSE, which is rigorously established and proved for the first time. Moreover, a novel technique is proposed for dealing with the nonlinear term in the equation, which plays an essential role in the error estimation. Finally, the numerical results confirm well with the theoretical findings.

math.NA↗

arXiv2020

Uniform convergence and stability of linearized fourth-order conservative compact scheme for Benjamin-Bona-Mahony-Burgers' equation

In the paper, a newly developed three-point fourth-order compact operator is utilized to construct an efficient compact finite difference scheme for the Benjamin-Bona-Mahony-Burgers' (BBMB) equation. Detailed derivation is carried out based on the reduction order method together with a three-level linearized technique. The conservative invariant, boundedness and unique solvability are studied at length. The uniform convergence is proved by the technical energy argument with the optimal convergence order $\mathcal{O}(τ^2+h^4)$ in the sense of the maximum norm. The almost unconditional stability can be achieved based on the uniform boundedness of the numerical solution. The present scheme is very efficient in practical computation since only a system of linear equations with a symmetric circulant matrix needing to be solved at each time step. The extensive numerical examples verify our theoretical results and demonstrate the superiority of the scheme when compared with state-of-the-art those in the references.

math.NA↗

arXiv2020

Multi-Task Reinforcement Learning based Mobile Manipulation Control for Dynamic Object Tracking and Grasping

Agile control of mobile manipulator is challenging because of the high complexity coupled by the robotic system and the unstructured working environment. Tracking and grasping a dynamic object with a random trajectory is even harder. In this paper, a multi-task reinforcement learning-based mobile manipulation control framework is proposed to achieve general dynamic object tracking and grasping. Several basic types of dynamic trajectories are chosen as the task training set. To improve the policy generalization in practice, random noise and dynamics randomization are introduced during the training process. Extensive experiments show that our policy trained can adapt to unseen random dynamic trajectories with about 0.1m tracking error and 75\% grasping success rate of dynamic objects. The trained policy can also be successfully deployed on a real mobile manipulator.

cs.RO↗

arXiv2020

Global bifurcation and stability of steady states for a bacterial colony model with density-suppressed motility

We investigate the structure and stability of the steady states for a bacterial colony model with density-suppressed motility. We treat the growth rate of bacteria as a bifurcation parameter to explore the local and global structure of the steady states. Relying on asymptotic analysis and the theory of Fredholm solvability, we derive the second-order approximate expression of the steady states. We analytically establish the stability criterion of the bifurcation solutions, and show that sufficiently large growth rate of bacteria leads to a stable uniform steady state. While the growth rate of bacteria is less than some certain value, there is pattern formation with the admissible wave mode. All the analytical results are corroborated by numerical simulations from different stages.

math.AP↗

arXiv2018

Global stability of traveling waves with oscillations for Nicholson's blowflies equation

For Nicholson's blowflies equation, a kind of reaction-diffusion equations with time-delay, when the ratio of birth rate coefficient and death rate coefficient satisfies $\frac{p}δ>e$, the large time-delay $r>0$ usually causes the traveling waves to be oscillatory. In this paper, we are interested in the global stability of these oscillatory traveling waves, in particular, the challenging case of the critical traveling waves with oscillations. We prove that, the critical oscillatory traveling waves are globally stable with the algebraic convergence rate $t^{-1/2}$, and the non-critical traveling waves are globally stable with the exponential convergence rate $t^{-1/2}e^{-μt}$ for a positive constant $μ$, where the initial perturbations around the oscillatory traveling wave in a weighted Sobolev can be arbitrarily large. The approach adopted is the technical weighted energy method with some new development in establishing the boundedness estimate of the oscillating solutions, which, with the help of optimal decay estimates by deriving the fundamental solutions for the linearized equations, can allow us to prove the global stability and to obtain the optimal convergence rates.

math.AP↗