arXiv subjects

Kaiwen Zhang

Publications and source records attributed to Kaiwen Zhang.

At least 37 records · Page 2Linked to original sources

arXiv2025

An Automatic Detection Method for Hematoma Features in Placental Abruption Ultrasound Images Based on Few-Shot Learning

Placental abruption is a severe complication during pregnancy, and its early accurate diagnosis is crucial for ensuring maternal and fetal safety. Traditional ultrasound diagnostic methods heavily rely on physician experience, leading to issues such as subjective bias and diagnostic inconsistencies. This paper proposes an improved model, EH-YOLOv11n (Enhanced Hemorrhage-YOLOv11n), based on small-sample learning, aiming to achieve automatic detection of hematoma features in placental ultrasound images. The model enhances performance through multidimensional optimization: it integrates wavelet convolution and coordinate convolution to strengthen frequency and spatial feature extraction; incorporates a cascaded group attention mechanism to suppress ultrasound artifacts and occlusion interference, thereby improving bounding box localization accuracy. Experimental results demonstrate a detection accuracy of 78%, representing a 2.5% improvement over YOLOv11n and a 13.7% increase over YOLOv8. The model exhibits significant superiority in precision-recall curves, confidence scores, and occlusion scenarios. Combining high accuracy with real-time processing, this model provides a reliable solution for computer-aided diagnosis of placental abruption, holding significant clinical application value.

cs.CV↗

arXiv2025

Conditional McKean-Vlasov control

Conditional McKean-Vlasov control problems involve controlling McKean-Vlasov diffusions where the interaction occurs through the law of the state process conditionally on it staying in a domain. Introduced by Lions in his 2016 lectures at the Collège de France, these problems have notable applications, particularly in systemic risk. We establish well-posedness and provide a general characterization of optimal controls using a new Pontryagin maximum principle in the probabilistic weak formulation. Unlike the classical approach based on forward-backward systems, our results connect the control problem to a generalized McKean-Vlasov backward stochastic differential equation (BSDE). We illustrate our framework with two applications: a version of the Schrödinger problem with killing, and a construction of equilibria in potential mean field games via McKean-Vlasov control.

math.PR↗

arXiv2025

Petri Nets-based Methods on Automatically Detecting for Concurrency Bugs in Rust Programs

Rust's memory safety guarantees, notably ownership and lifetime systems, have driven its widespread adoption. Concurrency bugs still occur in Rust programs, and existing detection approaches exhibit significant limitations: static analyzers suffer from context insensitivity and high false positives, while dynamic methods incur prohibitive runtime costs due to exponential path exploration. This paper presents a Petri net-based method for efficient, precise detection of Rust concurrency bugs. The method rests on three pillars: (1) A syntax-preserving program-to-Petri-net transformation tailored for target bug classes; (2) Semantics-preserving state compression via context-aware slicing; (3) Bug detection through efficient Petri net reachability analysis. The core innovation is its rigorous, control-flow-driven modeling of Rust's ownership semantics and synchronization primitives within the Petri net structure, with data operations represented as token movements. Integrated pointer analysis automates alias identification during transformation. Experiments on standard Rust concurrency benchmarks demonstrate that our method outperforms the state-of-the-art methods LockBud and Miri that are both tools of detecting concurrency bugs of Rust programs. Compared to LockBud, our approach reduces false positives by 35.7\% and false negatives by 28.3\% , which is obtained through our precise flow-sensitive pointer analysis. Compared with Miri that is a dynamic analysis tool, although Miri can obtain the same detection results, our method achieves 100% faster verification speed since our method takes a state reduce algorithm.

cs.PL↗

arXiv2025

V-ZOR: Enabling Verifiable Cross-Blockchain Communication via Quantum-Driven ZKP Oracle Relays

Cross-chain bridges and oracle DAOs represent some of the most vulnerable components of decentralized systems, with more than $2.8 billion lost due to trust failures, opaque validation behavior, and weak incentives. Current oracle designs are based on multisigs, optimistic assumptions, or centralized aggregation, exposing them to attacks and delays. Moreover, predictable committee selection enables manipulation, which threatens data integrity across chains. We propose V-ZOR, a verifiable oracle relay that integrates zero-knowledge proofs, quantum-grade randomness, and cross-chain restaking to mitigate these risks. Each oracle packet includes a Halo 2 proof verifying that the reported data was correctly aggregated using a deterministic median. To prevent committee manipulation, VZOR reseeds its VRF using auditable quantum entropy, ensuring unpredictable and secure selection of reporters. Reporters stake once on a shared restaking hub; any connected chain can submit a fraud proof to trigger slashing, removing the need for multisigs or optimistic assumptions. A prototype in Sepolia and Scroll achieves sub-300k gas verification, one-block latency, and a 10x increase in collusion cost. V-ZOR demonstrates that combining ZK attestation with quantum-randomized restaking enables a trust-minimized, high-performance oracle layer for cross-chain DeFi.

quant-ph↗

arXiv2025

A Range-Based Sharding (RBS) Protocol for Scalable Enterprise Blockchain

Blockchain technology offers decentralization and security but struggles with scalability, particularly in enterprise settings where efficiency and controlled access are paramount. Sharding is a promising solution for private blockchains, yet existing approaches face challenges in coordinating shards, ensuring fault tolerance with limited nodes, and minimizing the high overhead of consensus mechanisms like PBFT. This paper proposes the Range-Based Sharding (RBS) Protocol, a novel sharding mechanism tailored for enterprise blockchains, implemented on Quorum. Unlike traditional sharding models such as OmniLedger and non-sharding Corda framework, RBS employs a commit-reveal scheme for secure and unbiased shard allocation, ensuring fair validator distribution while reducing cross-shard transaction delays. Our approach enhances scalability by balancing computational loads across shards, reducing consensus overhead, and improving parallel transaction execution. Experimental evaluations demonstrate that RBS achieves significantly higher throughput and lower latency compared to existing enterprise sharding frameworks, making it a viable and efficient solution for largescale blockchain deployments.

cs.CR↗

arXiv2025

A Survey: Learning Embodied Intelligence from Physical Simulators and World Models

The pursuit of artificial general intelligence (AGI) has placed embodied intelligence at the forefront of robotics research. Embodied intelligence focuses on agents capable of perceiving, reasoning, and acting within the physical world. Achieving robust embodied intelligence requires not only advanced perception and control, but also the ability to ground abstract cognition in real-world interactions. Two foundational technologies, physical simulators and world models, have emerged as critical enablers in this quest. Physical simulators provide controlled, high-fidelity environments for training and evaluating robotic agents, allowing safe and efficient development of complex behaviors. In contrast, world models empower robots with internal representations of their surroundings, enabling predictive planning and adaptive decision-making beyond direct sensory input. This survey systematically reviews recent advances in learning embodied AI through the integration of physical simulators and world models. We analyze their complementary roles in enhancing autonomy, adaptability, and generalization in intelligent robots, and discuss the interplay between external simulation and internal modeling in bridging the gap between simulated training and real-world deployment. By synthesizing current progress and identifying open challenges, this survey aims to provide a comprehensive perspective on the path toward more capable and generalizable embodied AI systems. We also maintain an active repository that contains up-to-date literature and open-source projects at https://github.com/NJU3DV-LoongGroup/Embodied-World-Models-Survey.

cs.RO↗

arXiv2025

Deep Learning Based Concurrency Bug Detection and Localization

Concurrency bugs, caused by improper synchronization of shared resources in multi-threaded or distributed systems, are notoriously hard to detect and thus compromise software reliability and security. The existing deep learning methods face three main limitations. First, there is an absence of large and dedicated datasets of diverse concurrency bugs for them. Second, they lack sufficient representation of concurrency semantics. Third, binary classification results fail to provide finer-grained debug information such as precise bug lines. To address these problems, we propose a novel method for effective concurrency bug detection as well as localization. We construct a dedicated concurrency bug dataset to facilitate model training and evaluation. We then integrate a pre-trained model with a heterogeneous graph neural network (GNN), by incorporating a new Concurrency-Aware Code Property Graph (CCPG) that concisely and effectively characterizes concurrency semantics. To further facilitate debugging, we employ SubgraphX, a GNN-based interpretability method, which explores the graphs to precisely localize concurrency bugs, mapping them to specific lines of source code. On average, our method demonstrates an improvement of 10\% in accuracy and precision and 26\% in recall compared to state-of-the-art methods across diverse evaluation settings.

cs.SE↗

arXiv2025

EVMx: An FPGA-Based Smart Contract Processing Unit

Ethereum blockchain uses smart contracts (SCs) to implement decentralized applications (dApps). SCs are executed by the Ethereum virtual machine (EVM) running within an Ethereum client. Moreover, the EVM has been widely adopted by other blockchain platforms, including Solana, Cardano, Avalanche, Polkadot, and more. However, the EVM performance is limited by the constraints of the general-purpose computer it operates on. This work proposes offloading SC execution onto a dedicated hardware-based EVM. Specifically, EVMx is an FPGA-based SC execution engine that benefits from the inherent parallelism and high-speed processing capabilities of a hardware architecture. Synthesis results demonstrate a reduction in execution time of 61% to 99% for commonly used operation codes compared to CPU-based SC execution environments. Moreover, the execution time of Ethereum blocks on EVMx is up to 6x faster compared to analogous works in the literature. These results highlight the potential of the proposed architecture to accelerate SC execution and enhance the performance of EVM-compatible blockchains.

eess.SP↗

arXiv2025

NeurIT: Pushing the Limit of Neural Inertial Tracking for Indoor Robotic IoT

Inertial tracking is vital for robotic IoT and has gained popularity thanks to the ubiquity of low-cost inertial measurement units and deep learning-powered tracking algorithms. Existing works, however, have not fully utilized IMU measurements, particularly magnetometers, nor have they maximized the potential of deep learning to achieve the desired accuracy. To address these limitations, we introduce NeurIT, which elevates tracking accuracy to a new level. NeurIT employs a Time-Frequency Block-recurrent Transformer (TF-BRT) at its core, combining both RNN and Transformer to learn representative features in both time and frequency domains. To fully utilize IMU information, we strategically employ body-frame differentiation of magnetometers, considerably reducing the tracking error. We implement NeurIT on a customized robotic platform and conduct evaluation in various indoor environments. Experimental results demonstrate that NeurIT achieves a mere 1-meter tracking error over a 300-meter distance. Notably, it significantly outperforms state-of-the-art baselines by 48.21% on unseen data. Moreover, NeurIT demonstrates robustness in large urban complexes and performs comparably to the visual-inertial approach (Tango Phone) in vision-favored conditions while surpassing it in feature-sparse settings. We believe NeurIT takes an important step forward toward practical neural inertial tracking for ubiquitous and scalable tracking of robotic things. NeurIT is open-sourced here: https://github.com/aiot-lab/NeurIT.

cs.RO↗

arXiv2025

Epona: Autoregressive Diffusion World Model for Autonomous Driving

Diffusion models have demonstrated exceptional visual quality in video generation, making them promising for autonomous driving world modeling. However, existing video diffusion-based world models struggle with flexible-length, long-horizon predictions and integrating trajectory planning. This is because conventional video diffusion models rely on global joint distribution modeling of fixed-length frame sequences rather than sequentially constructing localized distributions at each timestep. In this work, we propose Epona, an autoregressive diffusion world model that enables localized spatiotemporal distribution modeling through two key innovations: 1) Decoupled spatiotemporal factorization that separates temporal dynamics modeling from fine-grained future world generation, and 2) Modular trajectory and video prediction that seamlessly integrate motion planning with visual modeling in an end-to-end framework. Our architecture enables high-resolution, long-duration generation while introducing a novel chain-of-forward training strategy to address error accumulation in autoregressive loops. Experimental results demonstrate state-of-the-art performance with 7.4\% FVD improvement and minutes longer prediction duration compared to prior works. The learned world model further serves as a real-time motion planner, outperforming strong end-to-end planners on NAVSIM benchmarks. Code will be publicly available at \href{https://github.com/Kevin-thu/Epona/}{https://github.com/Kevin-thu/Epona/}.

cs.CV↗

arXiv2024

WiP: Towards a Secure SECP256K1 for Crypto Wallets: Hardware Architecture and Implementation

The SECP256K1 elliptic curve algorithm is fundamental in cryptocurrency wallets for generating secure public keys from private keys, thereby ensuring the protection and ownership of blockchain-based digital assets. However, the literature highlights several successful side-channel attacks on hardware wallets that exploit SECP256K1 to extract private keys. This work proposes a novel hardware architecture for SECP256K1, optimized for side-channel attack resistance and efficient resource utilization. The architecture incorporates complete addition formulas, temporary registers, and parallel processing techniques, making elliptic curve point addition and doubling operations indistinguishable. Implementation results demonstrate an average reduction of 45% in LUT usage compared to similar works, emphasizing the design's resource efficiency.

cs.CR↗

arXiv2024

Maximal Extractable Value Mitigation Approaches in Ethereum and Layer-2 Chains: A Comprehensive Survey

Maximal Extractable Value (MEV) represents a pivotal challenge within the Ethereum ecosystem; it impacts the fairness, security, and efficiency of both Layer 1 (L1) and Layer 2 (L2) networks. MEV arises when miners or validators manipulate transaction ordering to extract additional value, often at the expense of other network participants. This not only affects user experience by introducing unpredictability and potential financial losses but also threatens the underlying principles of decentralization and trust. Given the growing complexity of blockchain applications, particularly with the increase of Decentralized Finance (DeFi) protocols, addressing MEV is crucial. This paper presents a comprehensive survey of MEV mitigation techniques as applied to both Ethereums L1 and various L2 solutions. We provide a novel categorization of mitigation strategies; we also describe the challenges, ranging from transaction sequencing and cryptographic methods to reconfiguring decentralized applications (DApps) to reduce front-running opportunities. We investigate their effectiveness, implementation challenges, and impact on network performance. By synthesizing current research, real-world applications, and emerging trends, this paper aims to provide a detailed roadmap for researchers, developers, and policymakers to understand and combat MEV in an evolving blockchain landscape.

cs.CR↗

arXiv2024

A Probabilistic Approach to Discounted Infinite Horizon and Invariant Mean Field Games

This paper considers discounted infinite horizon mean field games by extending the probabilistic weak formulation of the game as introduced by Carmona and Lacker (2015). Under similar assumptions as in the finite horizon game, we prove existence and uniqueness of solutions for the extended infinite horizon game. The key idea is to construct local versions of the previously considered stable topologies. Further, we analyze how sequences of finite horizon games approximate the infinite horizon one. Under a weakened Lasry-Lions monotonicity condition, we can quantify the convergence rate of solutions for the finite horizon games to the one for the infinite horizon game using a novel stability result for mean field games. Lastly, applying our results allows to solve the invariant mean field game as well.

math.OC↗

arXiv2024

Application of Kalman Filter in Stochastic Differential Equations

In areas such as finance, engineering, and science, we often face situations that change quickly and unpredictably. These situations are tough to handle and require special tools and methods capable of understanding and predicting what might happen next. Stochastic Differential Equations (SDEs) are renowned for modeling and analyzing real-world dynamical systems. However, obtaining the parameters, boundary conditions, and closed-form solutions of SDEs can often be challenging. In this paper, we will discuss the application of Kalman filtering theory to SDEs, including Extended Kalman filtering and Particle Extended Kalman filtering. We will explore how to fit existing SDE systems through filtering and track the original SDEs by fitting the obtained closed-form solutions. This approach aims to gather more information about these SDEs, which could be used in various ways, such as incorporating them into parameters of data-based SDE models.

eess.SY↗

arXiv2024

MultiConfederated Learning: Inclusive Non-IID Data handling with Decentralized Federated Learning

Federated Learning (FL) has emerged as a prominent privacy-preserving technique for enabling use cases like confidential clinical machine learning. FL operates by aggregating models trained by remote devices which owns the data. Thus, FL enables the training of powerful global models using crowd-sourced data from a large number of learners, without compromising their privacy. However, the aggregating server is a single point of failure when generating the global model. Moreover, the performance of the model suffers when the data is not independent and identically distributed (non-IID data) on all remote devices. This leads to vastly different models being aggregated, which can reduce the performance by as much as 50% in certain scenarios. In this paper, we seek to address the aforementioned issues while retaining the benefits of FL. We propose MultiConfederated Learning: a decentralized FL framework which is designed to handle non-IID data. Unlike traditional FL, MultiConfederated Learning will maintain multiple models in parallel (instead of a single global model) to help with convergence when the data is non-IID. With the help of transfer learning, learners can converge to fewer models. In order to increase adaptability, learners are allowed to choose which updates to aggregate from their peers.

cs.LG↗

arXiv2023

DiffMorpher: Unleashing the Capability of Diffusion Models for Image Morphing

Diffusion models have achieved remarkable image generation quality surpassing previous generative models. However, a notable limitation of diffusion models, in comparison to GANs, is their difficulty in smoothly interpolating between two image samples, due to their highly unstructured latent space. Such a smooth interpolation is intriguing as it naturally serves as a solution for the image morphing task with many applications. In this work, we present DiffMorpher, the first approach enabling smooth and natural image interpolation using diffusion models. Our key idea is to capture the semantics of the two images by fitting two LoRAs to them respectively, and interpolate between both the LoRA parameters and the latent noises to ensure a smooth semantic transition, where correspondence automatically emerges without the need for annotation. In addition, we propose an attention interpolation and injection technique and a new sampling schedule to further enhance the smoothness between consecutive images. Extensive experiments demonstrate that DiffMorpher achieves starkly better image morphing effects than previous methods across a variety of object categories, bridging a critical functional gap that distinguished diffusion models from GANs.

cs.CV↗

arXiv2023

Direct Linearly-Polarised Electroluminescence from Perovskite Nanoplatelet Superlattices

Polarised light is critical for a wide range of applications, but is usually generated by filtering unpolarised light, which leads to significant energy losses and requires additional optics. Herein, the direct emission of linearly-polarised light is achieved from light-emitting diodes (LEDs) made of CsPbI3 perovskite nanoplatelet superlattices. Through use of solvents with different vapour pressures, the self-assembly of perovskite nanoplatelets is achieved to enable fine control over the orientation (either face-up or edge-up) and therefore the transition dipole moment. As a result of the highly-uniform alignment of the nanoplatelets, as well as their strong quantum and dielectric confinement, large exciton fine-structure splitting is achieved at the film level, leading to pure-red LEDs exhibiting a high degree of linear polarisation of 74.4% without any photonic structures. This work unveils the possibilities of perovskite nanoplatelets as a highly promising source of linearly-polarised electroluminescence, opening up the development of next-generation 3D displays and optical communications from this highly versatile, solution-processable system.

physics.optics↗

arXiv2023

Early Detection for Multiversion Concurrency Control Conflicts in Hyperledger Fabric

Hyperledger Fabric is a popular permissioned blockchain system that features a highly modular and extensible system for deploying permissioned blockchains which are expected to have a major effect on a wide range of sectors. Unlike traditional blockchain systems such as Bitcoin and Ethereum, Hyperledger Fabric uses the EOV model for transaction processing: the submitted transactions are executed by the endorsing peer, ordered and batched by the ordering services, and validated by the validating peers. Due to this EOV workflow, a well-documented issue that arises is the multi-version concurrency control conflict. This happens when two transactions try to writes and read the same key in the ledger at the same time. Existing solutions to address this problem includes eliminating blocks in favor of streaming transactions, repairing conflicts during the ordering phase, and automatically merging the conflicting transactions using CRDT (Conflict Free Replicated Data) techniques. In this paper, we propose a novel solution called Early Detection for MVCC Conflicts. Our solution detects the conflicting transactions at an early stage of the transaction execution instead of processing them until the validation phase to be aborted. The advantage of our solution is that it detects conflict as soon as possible to minimize the overhead of conflicting transaction on the network resulting in the reduction of the end-to-end transaction latency and the increase of the system's effective throughput. We have successfully implemented our solution in Hyperledger Fabric. We propose three different implementations which realize early detection. Our results show that our solutions all perform better than the baseline Fabric, with our best solution SyncMap which improves the goodput by up to 23% and reduces the latency by up to 80%.

cs.DC↗