arXiv subjects

Qiang Shen

Publications and source records attributed to Qiang Shen.

At least 19 recordsLinked to original sources

arXiv2026

PAMAE: Phase-Aware-MoE Action Experts Towards Reliable Flow-Matching Vision-Language-Action Policies

Reliable action generation for multi-stage robotic manipulation remains challenging for Vision-Language-Action (VLA) models. While existing flow-matching VLA policies offer strong multimodal grounding and generalization, they typically employ a single shared action expert, limiting their ability to capture phase-specific control patterns across distinct execution stages. We propose a plug-and-play Phase-Aware Mixture-of-Experts Action Module (PAMAE), as a step towards more reliable phase-consistent action generation. PAMAE replaces the original flow-matching action expert with a sparse expert mixture while preserving the pretrained VLA backbone. PAMAE introduces a phase-aware router that leverages execution-phase cues to allocate action generation across experts, supported by a lightweight phase prediction head and a routing alignment objective. To stabilize specialization, we adopt a two-stage training scheme that first warms up the expert module under the standard flow-matching loss and then optimizes phase-consistent routing under auxiliary supervision. On multi-stage manipulation simulation tasks, PAMAE improves task success by up to \textbf{9.2\%} over strong VLA baselines. Further ablations show that both phase-supervised routing and staged optimization are essential for the observed gains. Our results highlight phase-consistent expert allocation as an effective mechanism for improving the reliability and action quality of flow-matching VLA policies.

cs.RO↗

arXiv2026

PhysReflect-VLA: Physical Feasibility and Self-Reflective Regulation for Reliable Vision-Language-Action Policies

Long-horizon robotic manipulation is highly sensitive to physically infeasible transitions, contact-induced disturbances, and the lack of effective self-correction during execution. Although Vision-Language-Action (VLA) models provide strong task grounding through multimodal learning, they typically generate actions in a feed-forward manner without explicitly checking physical feasibility or diagnosing execution errors online. We present PhysReflect-VLA, a plug-and-play execution-time reliability framework that augments VLA policies with physical feasibility evaluation and structured self-reflection in a closed-loop control pipeline. A Feasibility Operator evaluates whether candidate actions induce dynamically consistent state transitions; an Action Explanation Operator verifies transition coherence; and an LLM-based Reflection Module analyzes state discrepancies to generate corrective guidance for subsequent actions. A two-stage training procedure stabilizes feasibility modeling and integrates reflection into the control loop. Experiments on multi-stage, contact-rich real-world manipulation tasks show consistent improvements in stage-wise stability and overall task success compared with representative VLA baselines with an average gain of 5.4\%. Ablation results further indicate that feasibility checking and reflection-based correction both contribute to improved execution robustness. These results highlight the importance of embedding physical consistency checks and online self-reflection for reliable long-horizon robotic manipulation.

cs.RO↗

arXiv2026

Unsupervised Hyperspectral Image Super-Resolution via Self-Supervised Modality Decoupling

Fusion-based hyperspectral image super-resolution aims to fuse low-resolution hyperspectral images (LR-HSIs) and high-resolution multispectral images (HR-MSIs) to reconstruct high spatial and high spectral resolution images. Current methods typically apply direct fusion from the two modalities without effective supervision, leading to an incomplete perception of deep modality-complementary information and a limited understanding of inter-modality correlations. To address these issues, we propose a simple yet effective solution for unsupervised HMIF, revealing that modality decoupling is key to improving fusion performance. Specifically, we propose an end-to-end self-supervised Modality-Decoupled Spatial-Spectral Fusion (MossFuse) framework that decouples shared and complementary information across modalities and aggregates a concise representation of both LR-HSIs and HR-MSIs to reduce modality redundancy. Also, we introduce the subspace clustering loss as a clear guide to decouple modality-shared features from modality-complementary ones. Systematic experiments over multiple datasets demonstrate that our simple and effective approach consistently outperforms the existing HMIF methods while requiring considerably fewer parameters with reduced inference time. The source source code is in \href{https://github.com/dusongcheng/MossFuse}{MossFuse}.

eess.IV↗

arXiv2026

Pansharpening for Thin-Cloud Contaminated Remote Sensing Images: A Unified Framework and Benchmark Dataset

Pansharpening under thin cloudy conditions is a practically significant yet rarely addressed task, challenged by simultaneous spatial resolution degradation and cloud-induced spectral distortions. Existing methods often address cloud removal and pansharpening sequentially, leading to cumulative errors and suboptimal performance due to the lack of joint degradation modeling. To address these challenges, we propose a Unified Pansharpening Model with Thin Cloud Removal (Pan-TCR), an end-to-end framework that integrates physical priors. Motivated by theoretical analysis in the frequency domain, we design a frequency-decoupled restoration (FDR) block that disentangles the restoration of multispectral image (MSI) features into amplitude and phase components, each guided by complementary degradation-robust prompts: the near-infrared (NIR) band amplitude for cloud-resilient restoration, and the panchromatic (PAN) phase for high-resolution structural enhancement. To ensure coherence between the two components, we further introduce an interactive inter-frequency consistency (IFC) module, enabling cross-modal refinement that enforces consistency and robustness across frequency cues. Furthermore, we introduce the first real-world thin-cloud contaminated pansharpening dataset (PanTCR-GF2), comprising paired clean and cloudy PAN-MSI images, to enable robust benchmarking under realistic conditions. Extensive experiments on real-world and synthetic datasets demonstrate the superiority and robustness of Pan-TCR, establishing a new benchmark for pansharpening under realistic atmospheric degradations.

cs.CV↗

arXiv2025

Super4DR: 4D Radar-centric Self-supervised Odometry and Gaussian-based Map Optimization

Conventional odometry and mapping methods using visual or LiDAR data often struggle under poor illumination and adverse weather conditions. Although 4D radar is suited for such environments, its sparse and noisy point clouds hinder accurate odometry estimation, while the radar maps suffer from obscure and incomplete structures. Thus, we propose Super4DR, a 4D radar-centric framework for learning-based odometry estimation and gaussian-based map optimization. First, we design a cluster-aware odometry network that incorporates object-level cues from the clustered radar points for inter-frame matching, alongside a hierarchical self-supervision mechanism to overcome outliers through spatio-temporal consistency, knowledge transfer, and feature contrast. Second, we propose using 3D gaussians as an intermediate representation, coupled with a radar-specific growth strategy, selective separation, and multi-view regularization, to recover blurry map areas and those undetected based on image texture. Experiments show that Super4DR achieves a 67% performance gain over prior self-supervised methods, nearly matches supervised odometry, and narrows the map quality disparity with LiDAR while enabling multi-modal image rendering.

cs.RO↗

arXiv2024

3C: Confidence-Guided Clustering and Contrastive Learning for Unsupervised Person Re-Identification

Unsupervised person re-identification (Re-ID) aims to learn a feature network with cross-camera retrieval capability in unlabelled datasets. Although the pseudo-label based methods have achieved great progress in Re-ID, their performance in the complex scenario still needs to sharpen up. In order to reduce potential misguidance, including feature bias, noise pseudo-labels and invalid hard samples, accumulated during the learning process, in this pa per, a confidence-guided clustering and contrastive learning (3C) framework is proposed for unsupervised person Re-ID. This 3C framework presents three confidence degrees. i) In the clustering stage, the confidence of the discrepancy between samples and clusters is proposed to implement a harmonic discrepancy clustering algorithm (HDC). ii) In the forward-propagation training stage, the confidence of the camera diversity of a cluster is evaluated via a novel camera information entropy (CIE). Then, the clusters with high CIE values will play leading roles in training the model. iii) In the back-propagation training stage, the confidence of the hard sample in each cluster is designed and further used in a confidence integrated harmonic discrepancy (CHD), to select the informative sample for updating the memory in contrastive learning. Extensive experiments on three popular Re-ID benchmarks demonstrate the superiority of the proposed framework. Particularly, the 3C framework achieves state-of-the-art results: 86.7%/94.7%, 45.3%/73.1% and 47.1%/90.6% in terms of mAP/Rank-1 accuracy on Market-1501, the com plex datasets MSMT17 and VeRi-776, respectively. Code is available at https://github.com/stone5265/3C-reid.

cs.CV↗

arXiv2024

Uncovering the Over-smoothing Challenge in Image Super-Resolution: Entropy-based Quantification and Contrastive Optimization

PSNR-oriented models are a critical class of super-resolution models with applications across various fields. However, these models tend to generate over-smoothed images, a problem that has been analyzed previously from the perspectives of models or loss functions, but without taking into account the impact of data properties. In this paper, we present a novel phenomenon that we term the center-oriented optimization (COO) problem, where a model's output converges towards the center point of similar high-resolution images, rather than towards the ground truth. We demonstrate that the strength of this problem is related to the uncertainty of data, which we quantify using entropy. We prove that as the entropy of high-resolution images increases, their center point will move further away from the clean image distribution, and the model will generate over-smoothed images. Implicitly optimizing the COO problem, perceptual-driven approaches such as perceptual loss, model structure optimization, or GAN-based methods can be viewed. We propose an explicit solution to the COO problem, called Detail Enhanced Contrastive Loss (DECLoss). DECLoss utilizes the clustering property of contrastive learning to directly reduce the variance of the potential high-resolution distribution and thereby decrease the entropy. We evaluate DECLoss on multiple super-resolution benchmarks and demonstrate that it improves the perceptual quality of PSNR-oriented models. Moreover, when applied to GAN-based methods, such as RaGAN, DECLoss helps to achieve state-of-the-art performance, such as 0.093 LPIPS with 24.51 PSNR on 4x downsampled Urban100, validating the effectiveness and generalization of our approach.

eess.IV↗

arXiv2023

RZCR: Zero-shot Character Recognition via Radical-based Reasoning

The long-tail effect is a common issue that limits the performance of deep learning models on real-world datasets. Character image datasets are also affected by such unbalanced data distribution due to differences in character usage frequency. Thus, current character recognition methods are limited when applied in the real world, especially for the categories in the tail that lack training samples, e.g., uncommon characters. In this paper, we propose a zero-shot character recognition framework via radical-based reasoning, called RZCR, to improve the recognition performance of few-sample character categories in the tail. Specifically, we exploit radicals, the graphical units of characters, by decomposing and reconstructing characters according to orthography. RZCR consists of a visual semantic fusion-based radical information extractor (RIE) and a knowledge graph character reasoner (KGR). RIE aims to recognize candidate radicals and their possible structural relations from character images in parallel. The results are then fed into KGR to recognize the target character by reasoning with a knowledge graph. We validate our method on multiple datasets, and RZCR shows promising experimental results, especially on few-sample character datasets.

cs.CV↗

arXiv2021

DCF-ASN: Coarse-to-fine Real-time Visual Tracking via Discriminative Correlation Filter and Attentional Siamese Network

Discriminative correlation filters (DCF) and siamese networks have achieved promising performance on visual tracking tasks thanks to their superior computational efficiency and reliable similarity metric learning, respectively. However, how to effectively take advantages of powerful deep networks, while maintaining the real-time response of DCF, remains a challenging problem. Embedding the cross-correlation operator as a separate layer into siamese networks is a popular choice to enhance the tracking accuracy. Being a key component of such a network, the correlation layer is updated online together with other parts of the network. Yet, when facing serious disturbance, fused trackers may still drift away from the target completely due to accumulated errors. To address these issues, we propose a coarse-to-fine tracking framework, which roughly infers the target state via an online-updating DCF module first and subsequently, finely locates the target through an offline-training asymmetric siamese network (ASN). Benefitting from the guidance of DCF and the learned channel weights obtained through exploiting the given ground-truth template, ASN refines feature representation and implements precise target localization. Systematic experiments on five popular tracking datasets demonstrate that the proposed DCF-ASN achieves the state-of-the-art performance while exhibiting good tracking efficiency.

cs.CV↗

arXiv2021

Equalized Recovery State Estimators for Linear Systems with Delayed and Missing Observations

This paper presents a dynamic state observer design for discrete-time linear time-varying systems that robustly achieves equalized recovery despite delayed or missing observations, where the set of all temporal patterns for the missing or delayed data is modeled by a finite-length language. By introducing a mapping of the language onto a reduced event-based language, we design a state estimator that adapts based on the history of available data at each step, and satisfies equalized recovery for all patterns in the reduced language. In contrast to existing equalized recovery estimators, the proposed design considers the equalized recovery level as a decision variable, which enables us to directly obtain the global minimum for the intermediate recovery level, resulting in improved estimation performance. Finally, we demonstrate the effectiveness of the proposed observer when compared to existing approaches using several illustrative examples.

eess.SY↗

arXiv2020

Hyperspectral Classification Based on Lightweight 3-D-CNN With Transfer Learning

Recently, hyperspectral image (HSI) classification approaches based on deep learning (DL) models have been proposed and shown promising performance. However, because of very limited available training samples and massive model parameters, DL methods may suffer from overfitting. In this paper, we propose an end-to-end 3-D lightweight convolutional neural network (CNN) (abbreviated as 3-D-LWNet) for limited samples-based HSI classification. Compared with conventional 3-D-CNN models, the proposed 3-D-LWNet has a deeper network structure, less parameters, and lower computation cost, resulting in better classification performance. To further alleviate the small sample problem, we also propose two transfer learning strategies: 1) cross-sensor strategy, in which we pretrain a 3-D model in the source HSI data sets containing a greater number of labeled samples and then transfer it to the target HSI data sets and 2) cross-modal strategy, in which we pretrain a 3-D model in the 2-D RGB image data sets containing a large number of samples and then transfer it to the target HSI data sets. In contrast to previous approaches, we do not impose restrictions over the source data sets, in which they do not have to be collected by the same sensors as the target data sets. Experiments on three public HSI data sets captured by different sensors demonstrate that our model achieves competitive performance for HSI classification compared to several state-of-the-art methods

cs.CV↗

arXiv2020

Dislocation-induced Y segregation at basal-prismatic interfaces in Mg

Solute segregation at twin boundaries in Mg has been widely investigated, yet this phenomenon has not been studied at the equally important basal-prismatic interfaces. To fill this critical gap, this work investigates the segregation behavior of Y at basal-prismatic interfaces with various structures using atomistic simulations. The calculated interfacial energies show that short coherent interfaces and long semi-coherent interfaces containing disconnections and dislocations are more energetically stable than disordered interfaces, which is supported by our experimental observations. The segregation energy of Y at these lowest energy basal-prismatic interfaces shows a clear correlation with the atomic hydrostatic stress, highlighting the importance of local compressive stresses for segregation. In addition, sites around dislocations at the semi-coherent basal-prismatic interfaces demonstrate lower segregation energy, indicating that local defects such as interfacial dislocations can further enhance the segregation. In its entirety, this study indicates that the segregation of solutes can be affected by a number of different aspects of the local structure at complex interfaces in Mg alloys.

cond-mat.mtrl-sci↗

arXiv2020

Robust Correlation Tracking via Multi-channel Fused Features and Reliable Response Map

Benefiting from its ability to efficiently learn how an object is changing, correlation filters have recently demonstrated excellent performance for rapidly tracking objects. Designing effective features and handling model drifts are two important aspects for online visual tracking. This paper tackles these challenges by proposing a robust correlation tracking algorithm (RCT) based on two ideas: First, we propose a method to fuse features in order to more naturally describe the gradient and color information of the tracked object, and introduce the fused features into a background aware correlation filter to obtain the response map. Second, we present a novel strategy to significantly reduce noise in the response map and therefore ease the problem of model drift. Systematic comparative evaluations performed over multiple tracking benchmarks demonstrate the efficacy of the proposed approach.

cs.CV↗

arXiv2020

Multi-Modal Subjective Context Modelling and Recognition

Applications like personal assistants need to be aware ofthe user's context, e.g., where they are, what they are doing, and with whom. Context information is usually inferred from sensor data, like GPS sensors and accelerometers on the user's smartphone. This prediction task is known as context recognition. A well-defined context model is fundamental for successful recognition. Existing models, however, have two major limitations. First, they focus on few aspects, like location or activity, meaning that recognition methods based onthem can only compute and leverage few inter-aspect correlations. Second, existing models typically assume that context is objective, whereas in most applications context is best viewed from the user's perspective. Neglecting these factors limits the usefulness of the context model and hinders recognition. We present a novel ontological context model that captures five dimensions, namely time, location, activity, social relations and object. Moreover, our model defines three levels of description(objective context, machine context and subjective context) that naturally support subjective annotations and reasoning.An initial context recognition experiment on real-world data hints at the promise of our model.

cs.AI↗

arXiv2020

Tractable Compositions of Discrete-Time Control Barrier Functions with Application to Lane Keeping and Obstacle Avoidance

This paper introduces control barrier functions for discrete-time systems, which can be shown to be necessary and sufficient for controlled invariance of a given set. Moreover, we propose nonlinear discrete-time control barrier functions for partially control affine systems that lead to controlled invariance conditions that are affine in the control input, leading to a tractable formulation that enables us to handle the safety optimal control problem for a broader range of applications with more complicated safety conditions than existing approaches. In addition, we develop mixed-integer formulations for basic and secondary Boolean compositions of multiple control barrier functions and further provide mixed-integer constraints for piecewise control barrier functions. Finally, we apply these discrete-time control barrier function tools to automotive safety problems of lane keeping and obstacle avoidance, which are shown to be effective in simulation.

eess.SY↗

arXiv2020

Incremental Affine Abstraction of Nonlinear Systems

In this paper, we propose an incremental abstraction method for dynamically over-approximating nonlinear systems in a bounded domain by solving a sequence of linear programs, resulting in a sequence of affine upper and lower hyperplanes with expanding operating regions. Although the affine abstraction problem can be solved offline using a single linear program, existing approaches suffer from a computation space complexity that grows exponentially with the state dimension. Hence, the motivation for incremental abstraction is to reduce the space complexity for high-dimensional systems, but at the cost of yielding potentially worse abstractions/overapproximations. Specifically, we start with an operating region that is a subregion of the state space and compute two affine hyperplanes that bracket the nonlinear function locally. Then, by incrementally expanding the operating region, we dynamically update the two affine hyperplanes such that we eventually yield hyperplanes that are guaranteed to over-approximate the nonlinear system over the entire domain. Finally, the effectiveness of the proposed approach is demonstrated using numerical examples of high-dimensional nonlinear systems.

math.OC↗

arXiv2020

Uncovering the influence of common nonmetal impurities on the stability and strength of a $Σ$5 (310) grain boundary in Cu

Impurities are often driven to segregate to grain boundaries, which can significantly alter a material's thermal stability and mechanical behavior. To provide a comprehensive picture of this issue, the influence of a wide variety of common nonmetal impurities (H, B, C, N, O, Si, P and S) incorporated during service or materials processing are studied using first-principles simulations, with a focus on identifying changes to the energetics and mechanical strength of a Cu $Σ$5 (310) grain boundary. Changes to the grain boundary energy are found to be closely correlated with the covalent radii of the impurities and the volumetric deformations of polyhedra at the interface. The strengthening energies of each impurity are evaluated as a function of covalent radius and electronegativity, followed by first-principles-based tensile tests on selected impurities. The strengthening of a B-doped grain boundary comes from an enhancement of the charge density among the adjacent Cu atoms, which improves the connection between the two grains. Alternatively, the detrimental effect of O results from the reduction of charge density between the Cu atoms. This work deepens the understanding of the possible beneficial and harmful effects of impurities on grain boundaries, providing a guide for materials processing studies.

cond-mat.mtrl-sci↗

arXiv2020

Combined effects of nonmetallic impurities and planned metallic dopants on grain boundary energy and strength

Most research on nanocrystalline alloys has been focused on planned doping of metals with other metallic elements, but nonmetallic impurities are also prevalent in the real world. In this work, we report on the combined effects of metallic dopants and nonmetallic impurities on grain boundary energy and strength using first-principles calculations, with a $Σ$5 (310) grain boundary in Cu chosen as a model system. We find a clear correlation between the grain boundary energy and the change in excess free volume of doped grain boundaries. A combination of a larger substitutional dopant and an interstitial impurity can fill the excess free volume more efficiently and further reduce the grain boundary energy. We also find that the strengthening effects of dopants and impurities are dominated by the electronic interactions between the host Cu atoms and the two types of dopant elements. For example, the significant competing effects of metal dopants such as Zr, Nb, and Mo with impurities on the grain boundary strength are uncovered from the density of states of the d electrons. As a whole, this work deepens the field's understanding of the interaction between metallic dopants and nonmetallic impurities on grain boundary properties, providing a guide for improving the thermal stability of materials while avoiding embrittling effects.

cond-mat.mtrl-sci↗