arXiv subjects

Hao Shu

Publications and source records attributed to Hao Shu.

At least 19 recordsLinked to original sources

arXiv2025

Exact Recovery of Non-Random Missing Multidimensional Time Series via Temporal Isometric Delay-Embedding Transform

Non-random missing data is a ubiquitous yet undertreated flaw in multidimensional time series, fundamentally threatening the reliability of data-driven analysis and decision-making. Pure low-rank tensor completion, as a classical data recovery method, falls short in handling non-random missingness, both methodologically and theoretically. Hankel-structured tensor completion models provide a feasible approach for recovering multidimensional time series with non-random missing patterns. However, most Hankel-based multidimensional data recovery methods both suffer from unclear sources of Hankel tensor low-rankness and lack an exact recovery theory for non-random missing data. To address these issues, we propose the temporal isometric delay-embedding transform, which constructs a Hankel tensor whose low-rankness is naturally induced by the smoothness and periodicity of the underlying time series. Leveraging this property, we develop the \textit{Low-Rank Tensor Completion with Temporal Isometric Delay-embedding Transform} (LRTC-TIDT) model, which characterizes the low-rank structure under the \textit{Tensor Singular Value Decomposition} (t-SVD) framework. Once the prescribed non-random sampling conditions and mild incoherence assumptions are satisfied, the proposed LRTC-TIDT model achieves exact recovery, as confirmed by simulation experiments under various non-random missing patterns. Furthermore, LRTC-TIDT consistently outperforms existing tensor-based methods across multiple real-world tasks, including network flow reconstruction, urban traffic estimation, and temperature field prediction. Our implementation is publicly available at https://github.com/HaoShu2000/LRTC-TIDT.

cs.LG↗

arXiv2025

From Device to Dynamics: An Iterative Architectural Framework for High-Performance Single-Photon Detection at Room Temperature

Photon detection is a cornerstone of quantum technology, traditionally regarded as a static device-level operation constrained by the intrinsic physical properties of single-photon detectors (SPDs). Consequently, high-performance detection has been heavily reliant on superconducting technologies, whose requirement for cryogenic temperatures imposes significant infrastructure burdens and limits scalable deployment. To circumvent these constraints, we propose the Enhanced Single-Photon Detection (ESPD) framework, which shifts the photon-detection paradigm from device-centric optimization to an integrated quantum-information-processing (QIP) task. By incorporating state preparation, controlled operations, projective measurements, and multi-copy decision analysis, we establish a nonlinear dynamical model that reformulates detection as an iteratively enhanced process. This architecture enables systematic performance upgrades through structural design rather than material modification, allowing high-performance detection with exclusively room-temperature hardware. Through analytical approximations, Monte Carlo analysis, and numerical simulations, we show that the ESPD dynamics converge to a high-performance basin of attraction even when initialized by low-performance SPDs. While physical realization requires further component integration efforts, this work establishes a rigorous theoretical foundation for enhancing detection via architectural QIP principles. It provides not only a blueprint for next-generation room-temperature photon detection but also a general methodology for transcending device-level constraints in broader quantum technologies.

quant-ph↗

arXiv2025

Learning-Enhanced Observer for Linear Time-Invariant Systems with Parametric Uncertainty

This work introduces a learning-enhanced observer (LEO) for linear time-invariant systems with uncertain dynamics. Rather than relying solely on nominal models, the proposed framework treats the system matrices as optimizable variables and refines them through gradient-based minimization of a steady-state output discrepancy loss. The resulting data-informed surrogate model enables the construction of an improved observer that effectively compensates for moderate parameter uncertainty while preserving the structure of classical designs. Extensive Monte Carlo studies across diverse system dimensions show systematic and statistically significant reductions, typically exceeding 15\%, in normalized estimation error for both open-loop and Luenberger observers. These results demonstrate that modern learning mechanisms can serve as a powerful complement to traditional observer design, yielding more accurate and robust state estimation in uncertain systems. Codes are available at https://github.com/Hao-B-Shu/LTI_LEO.

cs.LG↗

arXiv2025

Vacuum-Signal Detection and the Principle-Level Feasibility of Arbitrarily Long-Distance Repeaterless Quantum Communication

Practical repeaterless quantum communication (PRQC) is constrained by the divergence of quantum bit error rate (QBER) arising from the interplay between channel loss and single-photon detector (SPD) dark counts. As the channel transmission rate decays with distance, vacuum-signal-induced dark counts inevitably dominate detection events beyond a finite range, driving QBER toward 50\% and rendering PRQC infeasible. Here, a theoretical framework termed vacuum-signal detection (VSD) is established to address this limitation at the level of principle. By employing controlled operations together with multi-copy analysis, the VSD paradigm enables vacuum-induced detection events to be identified and filtered without disturbing the encoded messages. Consequently, the non-vacuum signal ratio (NVSR) of the accepted signals can be stabilized at a high value irrespective of channel attenuation, thereby suppressing the fundamental QBER of PRQC within the secure bounds at any communication distance. As a result, PRQC can, in principle, remain feasible over arbitrarily long distances. By providing a rigorous theoretical resolution of the fundamental QBER-induced distance limitation, this work clarifies the principle-level scalability of PRQC. Furthermore, the vacuum-filtering framework developed here and its introduction of multi-copy analysis may also be of interest in a broader class of loss-sensitive or detection-based quantum tasks.

quant-ph↗

arXiv2025

Edge-Boundary-Texture Loss: A Tri-Class Generalization of Weighted Binary Cross-Entropy for Enhanced Edge Detection

Edge detection (ED) remains a fundamental task in computer vision, yet its performance is often hindered by the ambiguous nature of non-edge pixels near object boundaries. The widely adopted Weighted Binary Cross-Entropy (WBCE) loss treats all non-edge pixels uniformly, overlooking the structural nuances around edges and often resulting in blurred predictions. In this paper, we propose the Edge-Boundary-Texture (EBT) loss, a novel objective that explicitly divides pixels into three categories, edge, boundary, and texture, and assigns each a distinct supervisory weight. This tri-class formulation enables more structured learning by guiding the model to focus on both edge precision and contextual boundary localization. We theoretically show that the EBT loss generalizes the WBCE loss, with the latter becoming a limit case. Extensive experiments across multiple benchmarks demonstrate the superiority of the EBT loss both quantitatively and perceptually. Furthermore, the consistent use of unified hyperparameters across all models and datasets, along with robustness to their moderate variations, indicates that the EBT loss requires minimal fine-tuning and is easily deployable in practice.

cs.CV↗

arXiv2025

Robust Tensor Completion via Gradient Tensor Nulclear L1-L2 Norm for Traffic Data Recovery

In real-world scenarios, spatiotemporal traffic data frequently experiences dual degradation from missing values and noise caused by sensor malfunctions and communication failures. Therefore, effective data recovery methods are essential to ensure the reliability of downstream data-driven applications. while classical tensor completion methods have been widely adopted, they are incapable of modeling noise, making them unsuitable for complex scenarios involving simultaneous data missingness and noise interference. Existing Robust Tensor Completion (RTC) approaches offer potential solutions by separately modeling the actual tensor data and noise. However, their effectiveness is often constrained by the over-relaxation of convex rank surrogates and the suboptimal utilization of local consistency, leading to inadequate model accuracy. To address these limitations, we first introduce the tensor L1-L2 norm, a novel non-convex tensor rank surrogate that functions as an effective low-rank representation tool. Leveraging an advanced feature fusion strategy, we further develop the gradient tensor L1-L2 norm by incorporating the tensor L1-L2 norm in the gradient domain. By integrating the gradient tensor nuclear L1-L2 norm into the RTC framework, we propose the Robust Tensor Completion via Gradient Tensor Nuclear L1-L2 Norm (RTC-GTNLN) model, which not only fully exploits both global low-rankness and local consistency without trade-off parameter, but also effectively handles the dual degradation challenges of missing data and noise in traffic data. Extensive experiments conducted on multiple real-world traffic datasets demonstrate that the RTC-GTNLN model consistently outperforms existing state-of-the-art methods in complex recovery scenarios involving simultaneous missing values and noise.

cs.LG↗

arXiv2025

Binarization-Aware Adjuster for Discrete Decision Learning with an Application to Edge Detection

Discrete decision tasks in machine learning exhibit a fundamental misalignment between training and inference: models are optimized with continuous-valued outputs but evaluated using discrete predictions. This misalignment arises from the discontinuity of discretization operations, which prevents decision behavior from being directly incorporated into gradient-based optimization. To address this issue, we propose a theoretically grounded framework termed the Binarization-Aware Adjuster (BAA), which embeds binarization characteristics into continuous optimization. The framework is built upon the Distance Weight Function (DWF), which modulates loss contributions according to prediction correctness and proximity to the decision threshold, thereby aligning optimization emphasis with decision-critical regions while remaining compatible with standard learning pipelines. We apply the proposed BAA framework to the edge detection (ED) task, a representative binary decision problem. Experimental results on representative models and datasets show that incorporating BAA into optimization leads to consistent performance improvements, supporting its effectiveness. Overall, this work establishes a principled approach for aligning continuous optimization with discrete decision behavior, with its effectiveness demonstrated in a concrete application setting.

cs.CV↗

arXiv2025

Hyperspectral Anomaly Detection Fused Unified Nonconvex Tensor Ring Factors Regularization

In recent years, tensor decomposition-based approaches for hyperspectral anomaly detection (HAD) have gained significant attention in the field of remote sensing. However, existing methods often fail to fully leverage both the global correlations and local smoothness of the background components in hyperspectral images (HSIs), which exist in both the spectral and spatial domains. This limitation results in suboptimal detection performance. To mitigate this critical issue, we put forward a novel HAD method named HAD-EUNTRFR, which incorporates an enhanced unified nonconvex tensor ring (TR) factors regularization. In the HAD-EUNTRFR framework, the raw HSIs are first decomposed into background and anomaly components. The TR decomposition is then employed to capture the spatial-spectral correlations within the background component. Additionally, we introduce a unified and efficient nonconvex regularizer, induced by tensor singular value decomposition (TSVD), to simultaneously encode the low-rankness and sparsity of the 3-D gradient TR factors into a unique concise form. The above characterization scheme enables the interpretable gradient TR factors to inherit the low-rankness and smoothness of the original background. To further enhance anomaly detection, we design a generalized nonconvex regularization term to exploit the group sparsity of the anomaly component. To solve the resulting doubly nonconvex model, we develop a highly efficient optimization algorithm based on the alternating direction method of multipliers (ADMM) framework. Experimental results on several benchmark datasets demonstrate that our proposed method outperforms existing state-of-the-art (SOTA) approaches in terms of detection accuracy.

cs.CV↗

arXiv2025

Guaranteed Multidimensional Time Series Prediction via Deterministic Tensor Completion Theory

In recent years, the prediction of multidimensional time series data has become increasingly important due to its wide-ranging applications. Tensor-based prediction methods have gained attention for their ability to preserve the inherent structure of such data. However, existing approaches, such as tensor autoregression and tensor decomposition, often have consistently failed to provide clear assertions regarding the number of samples that can be exactly predicted. While matrix-based methods using nuclear norms address this limitation, their reliance on matrices limits accuracy and increases computational costs when handling multidimensional data. To overcome these challenges, we reformulate multidimensional time series prediction as a deterministic tensor completion problem and propose a novel theoretical framework. Specifically, we develop a deterministic tensor completion theory and introduce the Temporal Convolutional Tensor Nuclear Norm (TCTNN) model. By convolving the multidimensional time series along the temporal dimension and applying the tensor nuclear norm, our approach identifies the maximum forecast horizon for exact predictions. Additionally, TCTNN achieves superior performance in prediction accuracy and computational efficiency compared to existing methods across diverse real-world datasets, including climate temperature, network flow, and traffic ride data. Our implementation is publicly available at https://github.com/HaoShu2000/TCTNN.

cs.LG↗

arXiv2025

Symmetrization Weighted Binary Cross-Entropy: Modeling Perceptual Asymmetry for Human-Consistent Neural Edge Detection

Edge detection (ED) is a fundamental perceptual process in computer vision, forming the structural basis for high-level reasoning tasks such as segmentation, recognition, and scene understanding. Despite substantial progress achieved by deep neural networks, most ED models attain high numerical accuracy but fail to produce visually sharp and perceptually consistent edges, thereby limiting their reliability in intelligent vision systems. To address this issue, this study introduces the Symmetrization Weighted Binary Cross-Entropy (SWBCE) loss, a perception-inspired formulation that extends the conventional WBCE by incorporating prediction-guided symmetry. SWBCE explicitly models the perceptual asymmetry in human edge recognition, wherein edge decisions require stronger evidence than non-edge ones, aligning the optimization process with human perceptual discrimination. The resulting symmetric learning mechanism jointly enhances edge recall and suppresses false positives, achieving a superior balance between quantitative accuracy and perceptual fidelity. Extensive experiments across multiple benchmark datasets and representative ED architectures demonstrate that SWBCE can outperform existing loss functions in both numerical evaluation and visual quality. Particularly with the HED-EES model, the SSIM can be improved by about 15% on BRIND, and in all experiments, training by SWBCE consistently obtains the best perceptual results. Beyond edge detection, the proposed perceptual loss offers a generalizable optimization principle for soft computing and neural learning systems, particularly in scenarios where asymmetric perceptual reasoning plays a critical role.

cs.CV↗

arXiv2025

Boosting Edge Detection with Pixel-wise Feature Selection: The Extractor-Selector Paradigm

Deep learning has significantly advanced image edge detection (ED), primarily through improved feature extraction. However, most existing ED models apply uniform feature fusion across all pixels, ignoring critical differences between regions such as edges and textures. To address this limitation, we propose the Extractor-Selector (E-S) paradigm, a novel framework that introduces pixel-wise feature selection for more adaptive and precise fusion. Unlike conventional image-level fusion that applies the same convolutional kernel to all pixels, our approach dynamically selects relevant features at each pixel, enabling more refined edge predictions. The E-S framework can be seamlessly integrated with existing ED models without architectural changes, delivering substantial performance gains. It can also be combined with enhanced feature extractors for further accuracy improvements. Extensive experiments across multiple benchmarks confirm that our method consistently outperforms baseline ED models. For instance, on the BIPED2 dataset, the proposed framework can achieve over 7$\%$ improvements in ODS and OIS, and 22$\%$ improvements in AP, demonstrating its effectiveness and superiority.

cs.CV↗

arXiv2024

Improving Generated and Retrieved Knowledge Combination Through Zero-shot Generation

Open-domain Question Answering (QA) has garnered substantial interest by combining the advantages of faithfully retrieved passages and relevant passages generated through Large Language Models (LLMs). However, there is a lack of definitive labels available to pair these sources of knowledge. In order to address this issue, we propose an unsupervised and simple framework called Bi-Reranking for Merging Generated and Retrieved Knowledge (BRMGR), which utilizes re-ranking methods for both retrieved passages and LLM-generated passages. We pair the two types of passages using two separate re-ranking methods and then combine them through greedy matching. We demonstrate that BRMGR is equivalent to employing a bipartite matching loss when assigning each retrieved passage with a corresponding LLM-generated passage. The application of our model yielded experimental results from three datasets, improving their performance by +1.7 and +1.6 on NQ and WebQ datasets, respectively, and obtaining comparable result on TriviaQA dataset when compared to competitive baselines.

cs.CL↗

arXiv2024

A Comprehensive Study for Edge Detection: Ideal-Prior Guidance Learning, High-Precision Architecture, and Better Distinguishability Evaluation Benchmark

Image edge detection (ED) requires specialized architectures, reliable supervision, and rigorous evaluation criteria to ensure accurate localization. In this work, we present a framework for high-precision ED that jointly addresses architectural design, data supervision, and evaluation consistency. We propose SDPED, a compact ED model built upon Cascaded Skipping Density Blocks (CSDB), motivated by a task-adaptive architectural transfer from image super-resolution. By re-engineering texture-oriented structures for ED, SDPED effectively differentiates textures from edges while preserving fine spatial precision. Extensive experiments on four benchmark datasets (BRIND, UDED, MDBD, and BIPED2) demonstrate consistent performance improvements, particularly in Average Precision (AP), with gains of up to 22.5% on MDBD and 11.8% on BIPED2. In addition, we introduce an ideal-prior guidance strategy that incorporates noiseless data into training by treating labels as noise-free samples, providing a practical means to mitigate the subjectivity and noise inherent in human annotations. To enable better distinguishability and resolution-independent evaluation, we further study the criterion for assessing ED models, which suggests a stricter 1-pixel error tolerance distance. Overall, this work offers a coherent solution for high-precision ED and provides insights applicable to precision-oriented modeling in low-level and soft-computing-based vision tasks. Codes can be found on https://github.com/Hao-B-Shu/SDPED.

cs.CV↗

arXiv2023

Solve single photon detector problems

Single photon detector(SPD) problems arise in most quantum tasks, especially for measuring states going through high-lost channels. They are particularly prominent in quantum key distribution(QKD), which could be the most significant application in quantum information theory. In recent years, QKD distance has been improved dramatically but is still restricted because the bit error rate(QBER) caused by SPD dark counts will be out of control as the distance increases. If this problem can be solved, QKD can be implemented over arbitrarily long distances. However, previous solutions often result in impractical requirements such as superconductors while they can only reduce the dark count rate to finite low levels. In this paper, we solve SPD problems with today's technologies only. Although it is the no-cloning theorem that prevents a state from being measured multiple times to obtain a more reliable result, we propose a scheme circumventing the no-cloning theorem in certain tasks to allow a single state to be employed several times. The scheme demonstrates that imperfect detectors can provide nearly perfect results, namely, the QBER caused by dark counts can be reduced to arbitrarily low while in the meantime, detective efficiency can be improved to arbitrarily high. Consequently, QKD distance is not limited by the imperfect SPD anymore and can be improved from hundreds of kilometers to thousands without high-technology detectors. Furthermore, similar schemes can be applied for reducing measurement errors or improving the performance of sources. Finally, it is worth noting that although the paper is mainly discussed in the context of QKD, our scheme is an independent scheme that could be employed in other protocols wherever SPD are employed.

quant-ph↗

arXiv2023

Projectivities of informationally complete measurements

The physical problem behind informationally complete (IC) measurements is determining an unknown state statistically by measurement outcomes, known as state tomography. It is of central importance in quantum information processing such as channel estimating, device testing, quantum key distribution, etc. However, constructing such measurements with good properties is a long-standing problem. In this work, we investigate projective realizations of IC measurements. Conditions of informational completeness are presented with proofs first. Then the projective realizations of IC measurements, including proposing the first general construction of minimal projective IC measurements (MPICM) in no prime power dimensional systems, as well as determining an unknown state in $C^{n}$ via a single projective measurement with some kinds of optimalities in a larger system, are investigated. Finally, The results can be extended to local state tomography. Some discussions on employing several kinds of optimalities are also provided.

quant-ph↗

arXiv2023

Asymptotically Optimal prepare-measure Quantum Key Distribution Protocol

Quantum key distribution (QKD) could be the most significant application of quantum information theory. In nearly four decades, although substantial QKD protocols are developed, the BB84 protocol and its variants are still the most researched ones. It is well-known that the secure bound of qubit error rate (QBER) of BB84 protocol is about 11$\%$ while it can be increased to 12.6$\%$ by six-state protocol. It would not be surprising that employing more basis could increase the bound. However, what is the optimal protocol, and how to analyze it? In this paper, investigations of asymptotically optimal QKD protocols are proposed. Precisely, We present an abstraction of prepare-measure QKD protocols and investigate two special cases which are optimal among all protocols coding by the same states. Our analysis demonstrates that the asymptotically optimal QBER bounds coding by orthogonal qubits are about 27.28$\%$ for both memory C-NOT attacks and memoryless C-NOT attacks while the bounds coding by non-orthogonal states in two mutually unbiased bases are about 22.73$\%$ for memory and 28.69$\%$ for memoryless C-NOT attacks. The protocols are idealized but might be asymptotically realized while their optimality indicates the ultimate potential of QKD protocols. Although the analysis only contains a special kind of attack, it provides a framework for investigating such protocols.

quant-ph↗

arXiv2023

Quantum key distribution over noisy channels by the testing state method

Quantum key distribution(QKD) might be the most famous application of quantum information theory. The idea of QKD is not difficult to understand but in practical implementations, many problems are needed to be solved, for example, the noise of the channels. Previous works usually discuss the estimate of the channels and employ error-correcting procedures, whose feasibility and efficiency depend on the strength of the noise, or assist with entanglement distillation procedures, which often result in a large consumption of states while not all states can be distilled. This paper aims to study QKD over noisy channels including Pauli noises, amplitude damping noises, phase damping noises, collective noises as well as mixtures of them, in any strength without distillations. We provide a method, called the testing state method, to implement QKD protocols without errors over arbitrarily strength noisy channels. The method can be viewed as an error-correcting procedure, and can also be employed for other tasks.

quant-ph↗

arXiv2023

Measurement-Device-Independenization of Quantum Key Distribution Protocols

Quantum key distribution(QKD) allows the legitimate partner to establish a secret key whose security only depends on physical laws. In recent years, research on QKD by employing insecure measurement devices, namely measurement-device-independent QKD (MDI-QKD) is increased. MDI-QKD removes all attacks on measurement devices and thus an untrusted third party can be employed for measuring. However, a weakness of previous MDI-QKD protocols is the need for joint measurements such as Bell measurements whose efficiency is low in practice. On the other hand, can all QKD protocols become measurement-device-independent remains a problem. In this paper, we present a scheme making prepare-measure QKD protocols become MDI-QKD protocols, called $'measurement-device-independenization'$, which does not need to employ joint measurements and could be efficiently implemented by weak coherence sources. The protocol might look like the detector-device-independent(DDI) protocols but it is also secure under the Trojan horse attack. To illustrate this, we investigate the photon-number-adding(PNA) attack and present a scheme, called $'photon-number-purification'$, which can also be employed to close loopholes for previous protocols such as DDI and plug-and-play ones.

quant-ph↗