arXiv subjects

Dong Jiang

Publications and source records attributed to Dong Jiang.

18 recordsLinked to original sources

arXiv2026

X-Rec Technical Report

Recent advances in generative modeling have reshaped recommender systems by formulating recommendation as a next-item generation problem. Existing retrieval approaches primarily follow two paradigms: user-to-item (U2I) methods represent user context using one or a few deterministic embeddings, which limits the ability to capture diverse and multi-mode interests, while semantic-ID-based autoregressive (SID-AR) methods model more expressive distributions but suffer from quantization errors and the low throughput of sequential decoding. To address these limitations, we propose X-Rec to directly learn the recommendation distribution in the continuous item embedding space through flow matching and generate embedding triggers for approximate nearest neighbor retrieval. X-Rec incorporates three key designs to make this formulation effective and efficient. First, we introduce anchor conditioning to decompose generation into coarse semantic-region selection and fine-grained refinement. Second, we adopt Riemannian flow matching to align generative trajectories with the hyperspherical geometry of item embeddings. Third, we design a late-interaction diffusion Transformer that restricts repeated velocity-field estimation to the final Transformer layer. On a streaming benchmark, X-Rec substantially outperforms U2I baselines, matches the retrieval quality of SID-AR methods, and delivers 3.46x higher inference throughput than SID-AR. X-Rec has also been deployed as a new retrieval source for a specific vertical content on TikTok, where two consecutive launches have yielded significant improvements in both vertical engagement (+4.1484%) and general engagement (+0.0111%).

cs.IR↗

arXiv2026

OrthoDiffusion: A Generalizable Multi-Task Diffusion Foundation Model for Musculoskeletal MRI Interpretation

Musculoskeletal disorders represent a significant global health burden and are a leading cause of disability worldwide. While MRI is essential for accurate diagnosis, its interpretation remains exceptionally challenging. Radiologists must identify multiple potential abnormalities within complex anatomical structures across different imaging planes, a process that requires significant expertise and is prone to variability. We developed OrthoDiffusion, a unified diffusion-based foundation model designed for multi-task musculoskeletal MRI interpretation. The framework utilizes three orientation-specific 3D diffusion models, pre-trained in a self-supervised manner on 15,948 unlabeled knee MRI scans, to learn robust anatomical features from sagittal, coronal, and axial views. These view-specific representations are integrated to support diverse clinical tasks, including anatomical segmentation and multi-label diagnosis. Our evaluation demonstrates that OrthoDiffusion achieves excellent performance in the segmentation of 11 knee structures and the detection of 8 knee abnormalities. The model exhibited remarkable robustness across different clinical centers and MRI field strengths, consistently outperforming traditional supervised models. Notably, in settings where labeled data was scarce, OrthoDiffusion maintained high diagnostic precision using only 10\% of training labels. Furthermore, the anatomical representations learned from knee imaging proved highly transferable to other joints, achieving strong diagnostic performance across 11 diseases of the ankle and shoulder. These findings suggest that diffusion-based foundation models can serve as a unified platform for multi-disease diagnosis and anatomical segmentation, potentially improving the efficiency and accuracy of musculoskeletal MRI interpretation in real-world clinical workflows.

cs.CV↗

arXiv2026

A multimodal vision foundation model for generalizable knee pathology

Musculoskeletal disorders represent a leading cause of global disability, creating an urgent demand for precise interpretation of medical imaging. Current artificial intelligence (AI) approaches in orthopedics predominantly rely on task-specific, supervised learning paradigms. These methods are inherently fragmented, require extensive annotated datasets, and often lack generalizability across different modalities and clinical scenarios. The development of foundation models in this field has been constrained by the scarcity of large-scale, curated, and open-source musculoskeletal datasets. To address these challenges, we introduce OrthoFoundation, a multimodal vision foundation model optimized for musculoskeletal pathology. We constructed a pre-training dataset of 1.2 million unlabeled knee X-ray and MRI images from internal and public databases. Utilizing a Dinov3 backbone, the model was trained via self-supervised contrastive learning to capture robust radiological representations. OrthoFoundation achieves state-of-the-art (SOTA) performance across 14 downstream tasks. It attained superior accuracy in X-ray osteoarthritis diagnosis and ranked first in MRI structural injury detection. The model demonstrated remarkable label efficiency, matching supervised baselines using only 50% of labeled data. Furthermore, despite being pre-trained on knee images, OrthoFoundation exhibited exceptional cross-anatomy generalization to the hip, shoulder, and ankle. OrthoFoundation represents a significant advancement toward general-purpose AI for musculoskeletal imaging. By learning fundamental, joint-agnostic radiological semantics from large-scale multimodal data, it overcomes the limitations of conventional models, which provides a robust framework for reducing annotation burdens and enhancing diagnostic accuracy in clinical practice.

cs.CV↗

arXiv2025

The Illusion of Clinical Reasoning: A Benchmark Reveals the Pervasive Gap in Vision-Language Models for Clinical Competency

Background: The rapid integration of foundation models into clinical practice and public health necessitates a rigorous evaluation of their true clinical reasoning capabilities beyond narrow examination success. Current benchmarks, typically based on medical licensing exams or curated vignettes, fail to capture the integrated, multimodal reasoning essential for real-world patient care. Methods: We developed the Bones and Joints (B&J) Benchmark, a comprehensive evaluation framework comprising 1,245 questions derived from real-world patient cases in orthopedics and sports medicine. This benchmark assesses models across 7 tasks that mirror the clinical reasoning pathway, including knowledge recall, text and image interpretation, diagnosis generation, treatment planning, and rationale provision. We evaluated eleven vision-language models (VLMs) and six large language models (LLMs), comparing their performance against expert-derived ground truth. Results: Our results demonstrate a pronounced performance gap between task types. While state-of-the-art models achieved high accuracy, exceeding 90%, on structured multiple-choice questions, their performance markedly declined on open-ended tasks requiring multimodal integration, with accuracy scarcely reaching 60%. VLMs demonstrated substantial limitations in interpreting medical images and frequently exhibited severe text-driven hallucinations, often ignoring contradictory visual evidence. Notably, models specifically fine-tuned for medical applications showed no consistent advantage over general-purpose counterparts. Conclusions: Current artificial intelligence models are not yet clinically competent for complex, multimodal reasoning. Their safe deployment should currently be limited to supportive, text-based roles. Future advancement in core clinical tasks awaits fundamental breakthroughs in multimodal integration and visual understanding.

cs.CV↗

arXiv2025

FoMo4Wheat: Toward reliable crop vision foundation models with globally curated data

Vision-driven field monitoring is central to digital agriculture, yet models built on general-domain pretrained backbones often fail to generalize across tasks, owing to the interaction of fine, variable canopy structures with fluctuating field conditions. We present FoMo4Wheat, one of the first crop-domain vision foundation model pretrained with self-supervision on ImAg4Wheat, the largest and most diverse wheat image dataset to date (2.5 million high-resolution images collected over a decade at 30 global sites, spanning >2,000 genotypes and >500 environmental conditions). This wheat-specific pretraining yields representations that are robust for wheat and transferable to other crops and weeds. Across ten in-field vision tasks at canopy and organ levels, FoMo4Wheat models consistently outperform state-of-the-art models pretrained on general-domain dataset. These results demonstrate the value of crop-specific foundation models for reliable in-field perception and chart a path toward a universal crop foundation model with cross-species and cross-task capabilities. FoMo4Wheat models and the ImAg4Wheat dataset are publicly available online: https://github.com/PheniX-Lab/FoMo4Wheat and https://huggingface.co/PheniX-Lab/FoMo4Wheat. The demonstration website is: https://fomo4wheat.phenix-lab.com/.

cs.CV↗

arXiv2025

Real-Time Bit-Level Encryption of Full High-Definition Video Without Diffusion

Despite the widespread adoption of Shannon's confusion-diffusion architecture in image encryption, the implementation of diffusion to sequentially establish inter-pixel dependencies for attaining plaintext sensitivity constrains algorithmic parallelism, while the execution of multiple rounds of diffusion operations to meet the required sensitivity metrics incurs excessive computational overhead. Consequently, the pursuit of plaintext sensitivity through diffusion operations is the primary factor limiting the computational efficiency and throughput of video encryption algorithms, rendering them inadequate to meet the demands of real-time encryption for high-resolution video. To address the performance limitation, this paper proposes a real-time video encryption protocol based on heterogeneous parallel computing, which incorporates the SHA-256 hashes of original frames as input, employs multiple CPU threads to concurrently generate encryption-related data, and deploys numerous GPU threads to simultaneously encrypt pixels. By leveraging the extreme input sensitivity of the SHA hash, the proposed protocol achieves the required plaintext sensitivity metrics with only a single round of confusion and XOR operations, significantly reducing computational overhead. Furthermore, through eliminating the reliance on diffusion, it realizes the allocation of a dedicated GPU thread for encrypting each pixel within every channel, effectively enhancing algorithm's parallelism. The experimental results demonstrate that our approach not only exhibits superior statistical properties and robust security but also achieving delay-free bit-level encryption for 1920$\times$1080 resolution (full high definition) video at 30 FPS, with an average encryption time of 25.84 ms on a server equipped with an Intel Xeon Gold 6226R CPU and an NVIDIA GeForce RTX 3090 GPU.

cs.CR↗

arXiv2024

Experimental Quantum Byzantine Agreement on a Three-User Quantum Network with Integrated Photonics

Quantum communication networks are crucial for both secure communication and cryptographic networked tasks. Building quantum communication networks in a scalable and cost-effective way is essential for their widespread adoption, among which a stable and miniaturized high-quality quantum light source is a key component. Here, we establish a complete polarization entanglement-based fully connected network, which features an ultrabright integrated Bragg reflection waveguide quantum source, managed by an untrusted service provider, and a streamlined polarization analysis module, which requires only one single-photon detector for each end user. We perform a continuously working quantum entanglement distribution and create correlated bit strings between users. Within the framework of one-time universal hashing, we provide the first experimental implementation of source-independent quantum digital signatures using imperfect keys circumventing the necessity for private amplification. More importantly, we further beat the 1/3 fault-tolerance bound in Byzantine agreement, achieving unconditional security without relying on sophisticated techniques. Our results offer an affordable and practical route for addressing consensus challenges within the emerging quantum network landscape.

quant-ph↗

arXiv2024

Multi-Agent Generative Adversarial Interactive Self-Imitation Learning for AUV Formation Control and Obstacle Avoidance

Multiple autonomous underwater vehicles (multi-AUV) can cooperatively accomplish tasks that a single AUV cannot complete. Recently, multi-agent reinforcement learning has been introduced to control of multi-AUV. However, designing efficient reward functions for various tasks of multi-AUV control is difficult or even impractical. Multi-agent generative adversarial imitation learning (MAGAIL) allows multi-AUV to learn from expert demonstration instead of pre-defined reward functions, but suffers from the deficiency of requiring optimal demonstrations and not surpassing provided expert demonstrations. This paper builds upon the MAGAIL algorithm by proposing multi-agent generative adversarial interactive self-imitation learning (MAGAISIL), which can facilitate AUVs to learn policies by gradually replacing the provided sub-optimal demonstrations with self-generated good trajectories selected by a human trainer. Our experimental results in a multi-AUV formation control and obstacle avoidance task on the Gazebo platform with AUV simulator of our lab show that AUVs trained via MAGAISIL can surpass the provided sub-optimal expert demonstrations and reach a performance close to or even better than MAGAIL with optimal demonstrations. Further results indicate that AUVs' policies trained via MAGAISIL can adapt to complex and different tasks as well as MAGAIL learning from optimal demonstrations.

cs.RO↗

arXiv2023

Natural Adversarial Patch Generation Method Based on Latent Diffusion Model

Recently, some research show that deep neural networks are vulnerable to the adversarial attacks, the well-trainned samples or patches could be used to trick the neural network detector or human visual perception. However, these adversarial patches, with their conspicuous and unusual patterns, lack camouflage and can easily raise suspicion in the real world. To solve this problem, this paper proposed a novel adversarial patch method called the Latent Diffusion Patch (LDP), in which, a pretrained encoder is first designed to compress the natural images into a feature space with key characteristics. Then trains the diffusion model using the above feature space. Finally, explore the latent space of the pretrained diffusion model using the image denoising technology. It polishes the patches and images through the powerful natural abilities of diffusion models, making them more acceptable to the human visual system. Experimental results, both digital and physical worlds, show that LDPs achieve a visual subjectivity score of 87.3%, while still maintaining effective attack capabilities.

cs.CV↗

arXiv2023

Coexistence of multiuser entanglement distribution and classical light in optical fiber network with a semiconductor chip

Building communication links among multiple users in a scalable and robust way is a key objective in achieving large-scale quantum networks. In realistic scenario, noise from the coexisting classical light is inevitable and can ultimately disrupt the entanglement. The previous significant fully connected multiuser entanglement distribution experiments are conducted using dark fiber links and there is no explicit relation between the entanglement degradations induced by classical noise and its error rate. Here we fabricate a semiconductor chip with a high figure-of-merit modal overlap to directly generate broadband polarization entanglement. Our monolithic source maintains polarization entanglement fidelity above 96% for 42 nm bandwidth with a brightness of 1.2*10^7 Hz/mW. We perform a continuously working quantum entanglement distribution among three users coexisting with classical light. Under finite-key analysis, we establish secure keys and enable images encryption as well as quantum secret sharing between users. Our work paves the way for practical multiparty quantum communication with integrated photonic architecture compatible with real-world fiber optical communication network.

quant-ph↗

arXiv2023

Real-time chaotic video encryption based on multithreaded parallel confusion and diffusion

Due to the strong correlation between adjacent pixels, most image encryption schemes perform multiple rounds of confusion and diffusion to protect the image against attacks. Such operations, however, are time-consuming, cannot meet the real-time requirements of video encryption. Existing works, therefore, realize video encryption by simplifying the encryption process or encrypting specific parts of video frames, which results in lower security compared to image encryption. To solve the problem, this paper proposes a real-time chaotic video encryption strategy based on multithreaded parallel confusion and diffusion. It takes a video as the input, splits the frame into subframes, creates a set of threads to simultaneously perform five rounds of confusion and diffusion operations on corresponding subframes, and efficiently outputs the encrypted frames. The encryption speed evaluation shows that our method significantly improves the confusion and diffusion speed, realizes real-time 480x480, 576x576, and 768x768 24FPS video encryption using Intel Core i5-1135G7, Intel Core i7-8700, and Intel Xeon Gold 6226R, respectively. The statistical and security analysis prove that the deployed cryptosystems have outstanding statistical properties, can resist attacks, channel noise, and data loss. Compared with existing works, to the best of our knowledge, the proposed strategy achieves the fastest encryption speed, and realizes the first real-time chaotic video encryption that reaches the security level of image encryption. In addition, it is suitable for many confusion, diffusion algorithms and can be easily deployed with both hardware and software.

cs.CR↗

arXiv2022

NTIRE 2022 Challenge on Super-Resolution and Quality Enhancement of Compressed Video: Dataset, Methods and Results

This paper reviews the NTIRE 2022 Challenge on Super-Resolution and Quality Enhancement of Compressed Video. In this challenge, we proposed the LDV 2.0 dataset, which includes the LDV dataset (240 videos) and 95 additional videos. This challenge includes three tracks. Track 1 aims at enhancing the videos compressed by HEVC at a fixed QP. Track 2 and Track 3 target both the super-resolution and quality enhancement of HEVC compressed video. They require x2 and x4 super-resolution, respectively. The three tracks totally attract more than 600 registrations. In the test phase, 8 teams, 8 teams and 12 teams submitted the final results to Tracks 1, 2 and 3, respectively. The proposed methods and solutions gauge the state-of-the-art of super-resolution and quality enhancement of compressed video. The proposed LDV 2.0 dataset is available at https://github.com/RenYang-home/LDV_dataset. The homepage of this challenge (including open-sourced codes) is at https://github.com/RenYang-home/NTIRE22_VEnh_SR.

cs.CV↗

arXiv2020

FSD-10: A Dataset for Competitive Sports Content Analysis

Action recognition is an important and challenging problem in video analysis. Although the past decade has witnessed progress in action recognition with the development of deep learning, such process has been slow in competitive sports content analysis. To promote the research on action recognition from competitive sports video clips, we introduce a Figure Skating Dataset (FSD-10) for finegrained sports content analysis. To this end, we collect 1484 clips from the worldwide figure skating championships in 2017-2018, which consist of 10 different actions in men/ladies programs. Each clip is at a rate of 30 frames per second with resolution 1080 $\times$ 720. These clips are then annotated by experts in type, grade of execution, skater info, .etc. To build a baseline for action recognition in figure skating, we evaluate state-of-the-art action recognition methods on FSD-10. Motivated by the idea that domain knowledge is of great concern in sports field, we propose a keyframe based temporal segment network (KTSN) for classification and achieve remarkable performance. Experimental results demonstrate that FSD-10 is an ideal dataset for benchmarking action recognition algorithms, as it requires to accurately extract action motions rather than action poses. We hope FSD-10, which is designed to have a large collection of finegrained actions, can serve as a new challenge to develop more robust and advanced action recognition models.

cs.CV↗

arXiv2020

100Mbps Reconciliation for Quantum Key Distribution Using a Single Graphics Processing Unit

An efficient error reconciliation scheme is important for post-processing of quantum key distribution (QKD). Recently, a multi-matrix low-density parity-check codes based reconciliation algorithm which can provide remarkable perspectives for high efficiency information reconciliation was proposed. This paper concerns the improvement of reconciliation performance. Multi-matrix algorithm is implemented and optimized on the graphics processing unit (GPU) to obtain high reconciliation throughput. Experimental results indicate that GPU-based algorithm can highly improve reconciliation throughput to an average 85.67 Mbps and a maximum 102.084 Mbps with typical code rate and efficiency. This is the best performance of reconciliation on GPU platform to our knowledge.

quant-ph↗

arXiv2020

Multi-matrix rate-compatible reconciliation for quantum key distribution

Key reconciliation of quantum key distribution (QKD) is the process of correcting errors caused by channel noise and eavesdropper to identify the keys of two legitimate users. Reconciliation efficiency is the most important figure for judging the quality of a reconciliation scheme. To improve reconciliation efficiency, rate-compatible technologies was proposed for key reconciliation, which is denoted as the single-matrix ratecompatible reconciliation (SRCR). In this paper, a recently suggested technique called multi-matrix reconciliation is introduced into SRCR, which is referred to as the multi-matrix rate-compatible reconciliation (MRCR), to further improve reconciliation efficiency and promote the throughput of SRCR. Simulation results show that MRCR we proposed outperforms SRCR in reconciliation efficiency and throughput.

quant-ph↗

arXiv2019

Self-Organizing Quantum Networks

As the inevitable development trend of quantum key distribution, quantum networks have attracted extensive attention, and many prototypes have been deployed over recent years. Existing quantum networks based on optical fibers or quantum satellites can realize the metro and even global communications. However, for some application scenarios that require emergency or temporary communications, such networks cannot meet the requirements of rapid deployment, low cost, and high mobility. To solve this problem, we introduce an important concept of classical networks, i.e., self-organization, to quantum networks, and give two simple network prototypes based on acquisition, tracking, and pointing systems. In these networks, the users only need to deploy the network nodes, which will rapidly, automatically, and adaptively organize and mange a quantum network. Our method expands the application scope of quantum networks, and gives a new approach for the design and implementation of quantum networks. It also provides users with a low-cost access network solution, and can be used to fundamentally solve the security problem of classical self-organizing networks.

quant-ph↗

arXiv2019

Multi-Matrix Post-Processing for Quantum Key Distribution

Post-processing is a significant step in quantum key distribution(QKD), which is used for correcting the quantum-channel noise errors and distilling identical corrected keys between two distant legitimate parties. Efficient error reconciliation protocol, which can lead to an increase in the secure key generation rate, is one of the main performance indicators of QKD setups. In this paper, we propose a multi-low-density parity-check codes based reconciliation scheme, which can provide remarkable perspectives for highly efficient information reconciliation. With testing our approach through data simulation, we show that the proposed scheme combining multi-syndrome-based error rate estimation allows a more accurate estimation about the error rate as compared with random sampling and single-syndrome estimation techniques before the error correction, as well as a significant increase in the efficiency of the procedure without compromising security and sacrificing reconciliation efficiency.

quant-ph↗

arXiv2019

Color Recognition for Rubik's Cube Robot

In this paper, we proposed three methods to solve color recognition of Rubik's cube, which includes one offline method and two online methods. Scatter balance \& extreme learning machine (SB-ELM), a offline method, is proposed to illustrate the efficiency of training based method. We also point out the conception of color drifting which indicates offline methods are always ineffectiveness and can not work well in continuous change circumstance. By contrast, dynamic weight label propagation is proposed for labeling blocks color by known center blocks color of Rubik's cube. Furthermore, weak label hierarchic propagation, another online method, is also proposed for unknown all color information but only utilizes weak label of center block in color recognition. We finally design a Rubik's cube robot and construct a dataset to illustrate the efficiency and effectiveness of our online methods and to indicate the ineffectiveness of offline method by color drifting in our dataset.

cs.CV↗