arXiv subjects

Jun Zhang

Publications and source records attributed to Jun Zhang.

At least 325 records · Page 18Linked to original sources

arXiv2025

Dependence of halo properties on central-satellite magnitude gaps through weak lensing measurements

The magnitude gap between the central and satellite galaxies encodes information about the mass accretion history of a dark matter halo, and serves as a useful observational probe for the mass distribution in a halo. In this work, we perform the first weak lensing test of the connections between the magnitude gap and the halo profile. We measure the halo profiles of isolated central galaxies (ICGs) selected primarily from the SDSS Main Galaxy Sample. Halo mass and concentration are inferred by fitting stacked lensing profiles in bins of central luminosity, $L_\mathrm{c}$, and the central-satellite magnitude gap, $L_\mathrm{gap}$. We detect dependence on the magnitude gap in both halo properties. The dependence is the strongest in the ICG luminosity range of $10^{10.3}<L_\mathrm{c}[h^{-2}L_\odot]\leq 10^{10.7}$, where halos with smaller gaps have higher masses and lower concentrations. When $10^{10.7} <L_c[h^{-2}L_\odot] \leq 10^{11.1}$, however, no significant gap dependence is detected. In the range of $10^{9.9}<L_\mathrm{c}[h^{-2}L_\odot] \leq 10^{10.3}$, a disordering of the gap dependence is marginally observable. We compare the observational results with predictions by two lightcone catalogs built from the Illustris TNG300 and the Millennium simulations. The gap dependence in the two mock samples show overall consistency with observations, but neither matches them in all $L_\mathrm{c}$ bins to a quantitative level. We also compare the significance of the gap dependence on halo mass and concentration and find that our measurement prefers gap dependence in both parameters, while the halo mass dependence is preferred over the concentration if only one of the two dependencies is allowed.

astro-ph.CO↗

arXiv2025

Seed LiveInterpret 2.0: End-to-end Simultaneous Speech-to-speech Translation with Your Voice

Simultaneous Interpretation (SI) represents one of the most daunting frontiers in the translation industry, with product-level automatic systems long plagued by intractable challenges: subpar transcription and translation quality, lack of real-time speech generation, multi-speaker confusion, and translated speech inflation, especially in long-form discourses. In this study, we introduce Seed-LiveInterpret 2.0, an end-to-end SI model that delivers high-fidelity, ultra-low-latency speech-to-speech generation with voice cloning capabilities. As a fully operational product-level solution, Seed-LiveInterpret 2.0 tackles these challenges head-on through our novel duplex speech-to-speech understanding-generating framework. Experimental results demonstrate that through large-scale pretraining and reinforcement learning, the model achieves a significantly better balance between translation accuracy and latency, validated by human interpreters to exceed 70% correctness in complex scenarios. Notably, Seed-LiveInterpret 2.0 outperforms commercial SI solutions by significant margins in translation quality, while slashing the average latency of cloned speech from nearly 10 seconds to a near-real-time 3 seconds, which is around a near 70% reduction that drastically enhances practical usability.

cs.CL↗

arXiv2025

Silicon single-photon detector achieving over 84% photon detection efficiency with flexible operation modes

Silicon single-photon detectors (Si SPDs) play a crucial role in detecting single photons in the visible spectrum. For various applications, photon detection efficiency (PDE) is the most critical characteristic for effectively collecting photons. Here, we present a Si SPD with a remarkable PDE of up to 84.4% at 785 nm, supporting multiple operation modes. We design and fabricate a thick-junction Si single-photon avalanche diode (SPAD) that enhances the avalanche probability through a backside-illumination structure, while minimizing noise through the design of a doping-compensated avalanche region. To maximize PDE, we implement a readout circuit with a 50 V quenching voltage, enabling operation in free-running, gating, or hybrid modes. The SPAD, along with its readout circuits and affiliated circuits, is integrated into a compact SPD module. In free-running mode, the module achieves a maximum PDE of 84.4%, with a dark count rate of 260 cps, and an afterpulse probability of 2.9% at 268 K. This work provides a practical solution for applications requiring ultra-high-efficiency Si SPD with multiple operation modes.

quant-ph↗

arXiv2025

MEGA: Memory-Efficient 4D Gaussian Splatting for Dynamic Scenes

4D Gaussian Splatting (4DGS) has recently emerged as a promising technique for capturing complex dynamic 3D scenes with high fidelity. It utilizes a 4D Gaussian representation and a GPU-friendly rasterizer, enabling rapid rendering speeds. Despite its advantages, 4DGS faces significant challenges, notably the requirement of millions of 4D Gaussians, each with extensive associated attributes, leading to substantial memory and storage cost. This paper introduces a memory-efficient framework for 4DGS. We streamline the color attribute by decomposing it into a per-Gaussian direct color component with only 3 parameters and a shared lightweight alternating current color predictor. This approach eliminates the need for spherical harmonics coefficients, which typically involve up to 144 parameters in classic 4DGS, thereby creating a memory-efficient 4D Gaussian representation. Furthermore, we introduce an entropy-constrained Gaussian deformation technique that uses a deformation field to expand the action range of each Gaussian and integrates an opacity-based entropy loss to limit the number of Gaussians, thus forcing our model to use as few Gaussians as possible to fit a dynamic scene well. With simple half-precision storage and zip compression, our framework achieves a storage reduction by approximately 190$\times$ and 125$\times$ on the Technicolor and Neural 3D Video datasets, respectively, compared to the original 4DGS. Meanwhile, it maintains comparable rendering speeds and scene representation quality, setting a new standard in the field. Code is available at https://github.com/Xinjie-Q/MEGA.

cs.CV↗

arXiv2025

AI and Deep Learning for Terahertz Ultra-Massive MIMO: From Model-Driven Approaches to Foundation Models

This study explored the transformative potential of artificial intelligence (AI) in addressing the challenges posed by terahertz ultra-massive multiple-input multiple-output (UM-MIMO) systems. It begins by outlining the characteristics of terahertz UM-MIMO systems and identifies three primary challenges for transceiver design: computational complexity, modeling difficulty, and measurement limitations. The study posits that AI provides a promising solution to these challenges. Three systematic research roadmaps are proposed for developing AI algorithms tailored to terahertz UM-MIMO systems. The first roadmap, model-driven deep learning (DL), emphasizes the importance of leveraging available domain knowledge and advocates the adoption of AI only to enhance bottleneck modules within an established signal processing or optimization framework. Four essential steps are discussed: algorithmic frameworks, basis algorithms, loss-function design, and neural architecture design. The second roadmap presents channel state information (CSI) foundation models, aimed at unifying the design of different transceiver modules by focusing on their shared foundation, that is, the wireless channel. The training of a single compact foundation model is proposed to estimate the score function of wireless channels, which serve as a versatile prior for designing a wide variety of transceiver modules. Four essential steps are outlined: general frameworks, conditioning, site-specific adaptation, joint design of CSI foundation models, and model-driven DL. The third roadmap aims to explore potential directions for applying pretrained large language models (LLMs) to terahertz UM-MIMO systems. Several application scenarios are envisioned, including LLM-based estimation, optimization, search, network management, and protocol understanding. Finally, the study highlights open problems and future research directions.

eess.SP↗

arXiv2025

Optimal Transceiver Design in Over-the-Air Federated Distillation

The rapid proliferation and growth of artificial intelligence (AI) has led to the development of federated learning (FL). FL allows wireless devices (WDs) to cooperatively learn by sharing only local model parameters, without needing to share the entire dataset. However, the emergence of large AI models has made existing FL approaches inefficient, due to the significant communication overhead required. In this paper, we propose a novel over-the-air federated distillation (FD) framework by synergizing the strength of FL and knowledge distillation to avoid the heavy local model transmission. Instead of sharing the model parameters, only the WDs' model outputs, referred to as knowledge, are shared and aggregated over-the-air by exploiting the superposition property of the multiple-access channel. We shall study the transceiver design in over-the-air FD, aiming to maximize the learning convergence rate while meeting the power constraints of the transceivers. The main challenge lies in the intractability of the learning performance analysis, as well as the non-convex nature and the optimization spanning the whole FD training period. To tackle this problem, we first derive an analytical expression of the convergence rate in over-the-air FD. Then, the closed-form optimal solutions of the WDs' transmit power and the estimator for over-the-air aggregation are obtained given the receiver combining strategy. Accordingly, we put forth an efficient approach to find the optimal receiver beamforming vector via semidefinite relaxation. We further prove that there is no optimality gap between the original and relaxed problem for the receiver beamforming design. Numerical results will show that the proposed over-the-air FD approach achieves a significant reduction in communication overhead, with only a minor compromise in testing accuracy compared to conventional FL benchmarks.

eess.SP↗

arXiv2025

Fusing Large Language Models with Temporal Transformers for Time Series Forecasting

Recently, large language models (LLMs) have demonstrated powerful capabilities in performing various tasks and thus are applied by recent studies to time series forecasting (TSF) tasks, which predict future values with the given historical time series. Existing LLM-based approaches transfer knowledge learned from text data to time series prediction using prompting or fine-tuning strategies. However, LLMs are proficient at reasoning over discrete tokens and semantic patterns but are not initially designed to model continuous numerical time series data. The gaps between text and time series data lead LLMs to achieve inferior performance to a vanilla Transformer model that is directly trained on TSF data. However, the vanilla Transformers often struggle to learn high-level semantic patterns. In this paper, we design a novel Transformer-based architecture that complementarily leverages LLMs and vanilla Transformers, so as to integrate the high-level semantic representations learned by LLMs into the temporal information encoded by time series Transformers, where a hybrid representation is obtained by fusing the representations from the LLM and the Transformer. The resulting fused representation contains both historical temporal dynamics and semantic variation patterns, allowing our model to predict more accurate future values. Experiments on benchmark datasets demonstrate the effectiveness of the proposed approach.

cs.CL↗

arXiv2025

OpenThinkIMG: Learning to Think with Images via Visual Tool Reinforcement Learning

While humans can flexibly leverage interactive visual cognition for complex problem-solving, enabling Large Vision-Language Models (LVLMs) to learn similarly adaptive behaviors with visual tools remains challenging. A significant hurdle is the current lack of standardized infrastructure, which hinders integrating diverse tools, generating rich interaction data, and training robust agents effectively. To address these gaps, we introduce OpenThinkIMG, the first open-source, comprehensive end-to-end framework for tool-augmented LVLMs. It features standardized vision tool interfaces, scalable trajectory generation for policy initialization, and a flexible training environment. Furthermore, considering supervised fine-tuning (SFT) on static demonstrations offers limited policy generalization for dynamic tool invocation, we propose a novel reinforcement learning (RL) framework V-ToolRL to train LVLMs to learn adaptive policies for invoking external vision tools. V-ToolRL enables LVLMs to autonomously discover optimal tool-usage strategies by directly optimizing for task success using feedback from tool interactions. We empirically validate V-ToolRL on challenging chart reasoning tasks. Our RL-trained agent, built upon a Qwen2-VL-2B, significantly outperforms its SFT-initialized counterpart (+28.83 points) and surpasses established supervised tool-learning baselines like Taco and CogCom by an average of +12.7 points. Notably, it also surpasses prominent closed-source models like GPT-4.1 by +8.68 accuracy points. We hope OpenThinkIMG can serve as a foundational framework for advancing dynamic, tool-augmented visual reasoning, helping the community develop AI agents that can genuinely "think with images".

cs.CV↗

arXiv2025

WiLLM: an Open Framework for LLM Services over Wireless Systems

Large Language Model (LLM) services fundamentally differ from traditional Deep Neural Network (DNN) applications in wireless networks. We identify three critical distinctions: (1) unlike traditional DNNs with unidirectional data flows, LLM's multimodal interactions create bidirectional heavy loads with contrasting bottlenecks, requiring direction-aware resource scheduling; (2) while traditional DNNs exhibit fixed computational patterns, LLM's highly variable inference times interact complexly with network slicing, causing dynamic bottleneck migration; and (3) in contrast to predictable DNN traffic, LLM's token streams demonstrate unprecedented burstiness and state dependencies. These insights motivate WiLLM, the first open-source framework, implemented as a wireless platform, for LLM service research. Built on OpenAirInterface, WiLLM introduces several technical innovations: dynamic slice compatibility, universal UE compatibility through application-layer tunneling, multi-UE multi-slice scheduling, dual-mode resource allocation, and cross-layer APIs. In addition, WiLLM eliminates the need for specialized wireless expertise, enabling researchers and developers to experiment with LLM services over realistic cellular networks. We demonstrate the platform's capabilities through a smart glasses case study and provide a comprehensive dataset of \~1.6 million synchronized measurements. The complete system, dataset, and appendix are available at https://openwillm.github.io.

cs.NI↗

arXiv2025

PaddleOCR 3.0 Technical Report

This technical report introduces PaddleOCR 3.0, an Apache-licensed open-source toolkit for OCR and document parsing. To address the growing demand for document understanding in the era of large language models, PaddleOCR 3.0 presents three major solutions: (1) PP-OCRv5 for multilingual text recognition, (2) PP-StructureV3 for hierarchical document parsing, and (3) PP-ChatOCRv4 for key information extraction. Compared to mainstream vision-language models (VLMs), these models with fewer than 100 million parameters achieve competitive accuracy and efficiency, rivaling billion-parameter VLMs. In addition to offering a high-quality OCR model library, PaddleOCR 3.0 provides efficient tools for training, inference, and deployment, supports heterogeneous hardware acceleration, and enables developers to easily build intelligent document applications.

cs.CV↗

arXiv2025

Force-IMU Fusion-Based Sensing Acupuncture Needle and Quantitative Analysis System for Acupuncture Manipulations

Acupuncture, one of the key therapeutic methods in Traditional Chinese Medicine (TCM), has been widely adopted in various clinical fields. Quantitative research on acupuncture manipulation parameters is critical to achieve standardized techniques. However, quantitative mechanical detection of acupuncture parameters remains limited. This study establishes a kinematic and dynamic model of acupuncture, identifying key parameters such as lifting-thrusting force, acceleration, velocity, displacement, as well as twirling-rotating angular velocity and angle. To measure these critical parameters, we propose a quantitative system comprising a sensing needle equipped with a force sensor and an inertial measurement unit (IMU), as well as an external camera module to capture image information. By fusing visual and IMU data, we accurately identify the stationary or motion states of the needle, enabling segmented computation of lifting-thrusting velocity and displacement. The experimental results demonstrate that the sensing needle achieves comprehensive detection with high precision, featuring a nonlinearity error of 0.45% in force measurement and an RMSE of 1.2 mm in displacement. The extracted parameters provide an objective description of the operational characteristics and motion patterns of the four basic acupuncture manipulations. These findings provide valuable tools and methods for research in acupuncture standardization.

eess.SY↗

arXiv2025

LVM4CSI: Enabling Direct Application of Pre-Trained Large Vision Models for Wireless Channel Tasks

Accurate channel state information (CSI) is critical to the performance of wireless communication systems, especially with the increasing scale and complexity introduced by 5G and future 6G technologies. While artificial intelligence (AI) offers a promising approach to CSI acquisition and utilization, existing methods largely depend on task-specific neural networks (NNs) that require expert-driven design and large training datasets, limiting their generalizability and practicality. To address these challenges, we propose LVM4CSI, a general and efficient framework that leverages the structural similarity between CSI and computer vision (CV) data to directly apply large vision models (LVMs) pre-trained on extensive CV datasets to wireless tasks without any fine-tuning, in contrast to large language model-based methods that generally necessitate fine-tuning. LVM4CSI maps CSI tasks to analogous CV tasks, transforms complex-valued CSI into visual formats compatible with LVMs, and integrates lightweight trainable layers to adapt extracted features to specific communication objectives. We validate LVM4CSI through three representative case studies, including channel estimation, human activity recognition, and user localization. Results demonstrate that LVM4CSI achieves comparable or superior performance to task-specific NNs, including an improvement exceeding 9.61 dB in channel estimation and approximately 40% reduction in localization error. Furthermore, it significantly reduces the number of trainable parameters and eliminates the need for task-specific NN design.

cs.IT↗

arXiv2025

Symbolic identification of tensor equations in multidimensional physical fields

Recently, data-driven methods have shown great promise for discovering governing equations from simulation or experimental data. However, most existing approaches are limited to scalar equations, with few capable of identifying tensor relationships. In this work, we propose a general data-driven framework for identifying tensor equations, referred to as Symbolic Identification of Tensor Equations (SITE). The core idea of SITE--representing tensor equations using a host-plasmid structure--is inspired by the multidimensional gene expression programming (M-GEP) approach. To improve the robustness of the evolutionary process, SITE adopts a genetic information retention strategy. Moreover, SITE introduces two key innovations beyond conventional evolutionary algorithms. First, it incorporates a dimensional homogeneity check to restrict the search space and eliminate physically invalid expressions. Second, it replaces traditional linear scaling with a tensor linear regression technique, greatly enhancing the efficiency of numerical coefficient optimization. We validate SITE using two benchmark scenarios, where it accurately recovers target equations from synthetic data, showing robustness to noise and small sample sizes. Furthermore, SITE is applied to identify constitutive relations directly from molecular simulation data, which are generated without reliance on macroscopic constitutive models. It adapts to both compressible and incompressible flow conditions and successfully identifies the corresponding macroscopic forms, highlighting its potential for data-driven discovery of tensor equation.

math-ph↗

arXiv2025

The Dependence of Dark Matter Halo Properties on the Morphology of Their Central Galaxies from Weak Lensing

Xu \& Jing reported a monotonic relationship between host halo mass $M_h$ and the morphology of massive central galaxies, characterized by the Sérsic index $n$, at fixed stellar mass, suggesting that morphology could serve as a good secondary proxy for halo mass. Since their results were derived using the indirect abundance matching method, we further investigate the connection between halo properties and central galaxy morphology using weak gravitational lensing. We apply galaxy-galaxy lensing to measure the excess surface density around CMASS central galaxies with stellar masses in the range of $11.3 < \log M_*/{\rm M_\odot} < 11.7$, using the HSC shear catalog processed through the Fourier\_Quad pipeline. By dividing the sample based on $n$, we confirm a positive correlation between $n$ and $M_h$, and observe a possible evidence of the positive correlation of $n$ and halo concentration. After accounting for color, we find that neither color nor morphology alone can determine halo mass, suggesting that a combination of both may serve as a better secondary proxy. In comparison to hydrodynamic simulations, we find that TNG300 produce much weaker correlations between $M_h$ and $n$. Furthermore, using SIMBA simulations with different feedback mode, we find jet-mode active galactic nuclei feedback might be related to the relationship of Sérsic index and halo mass.

astro-ph.GA↗

arXiv2025

Neural Graph Map: Dense Mapping with Efficient Loop Closure Integration

Neural field-based SLAM methods typically employ a single, monolithic field as their scene representation. This prevents efficient incorporation of loop closure constraints and limits scalability. To address these shortcomings, we propose a novel RGB-D neural mapping framework in which the scene is represented by a collection of lightweight neural fields which are dynamically anchored to the pose graph of a sparse visual SLAM system. Our approach shows the ability to integrate large-scale loop closures, while requiring only minimal reintegration. Furthermore, we verify the scalability of our approach by demonstrating successful building-scale mapping taking multiple loop closures into account during the optimization, and show that our method outperforms existing state-of-the-art approaches on large scenes in terms of quality and runtime. Our code is available open-source at https://github.com/KTH-RPL/neural_graph_mapping.

cs.CV↗

arXiv2025

ReconX: Reconstruct Any Scene from Sparse Views with Video Diffusion Model

Advancements in 3D scene reconstruction have transformed 2D images from the real world into 3D models, producing realistic 3D results from hundreds of input photos. Despite great success in dense-view reconstruction scenarios, rendering a detailed scene from insufficient captured views is still an ill-posed optimization problem, often resulting in artifacts and distortions in unseen areas. In this paper, we propose ReconX, a novel 3D scene reconstruction paradigm that reframes the ambiguous reconstruction challenge as a temporal generation task. The key insight is to unleash the strong generative prior of large pre-trained video diffusion models for sparse-view reconstruction. However, 3D view consistency struggles to be accurately preserved in directly generated video frames from pre-trained models. To address this, given limited input views, the proposed ReconX first constructs a global point cloud and encodes it into a contextual space as the 3D structure condition. Guided by the condition, the video diffusion model then synthesizes video frames that are both detail-preserved and exhibit a high degree of 3D consistency, ensuring the coherence of the scene from various perspectives. Finally, we recover the 3D scene from the generated video through a confidence-aware 3D Gaussian Splatting optimization scheme. Extensive experiments on various real-world datasets show the superiority of our ReconX over state-of-the-art methods in terms of quality and generalizability.

cs.CV↗

arXiv2025

From Virtual Games to Real-World Play

We introduce RealPlay, a neural network-based real-world game engine that enables interactive video generation from user control signals. Unlike prior works focused on game-style visuals, RealPlay aims to produce photorealistic, temporally consistent video sequences that resemble real-world footage. It operates in an interactive loop: users observe a generated scene, issue a control command, and receive a short video chunk in response. To enable such realistic and responsive generation, we address key challenges including iterative chunk-wise prediction for low-latency feedback, temporal consistency across iterations, and accurate control response. RealPlay is trained on a combination of labeled game data and unlabeled real-world videos, without requiring real-world action annotations. Notably, we observe two forms of generalization: (1) control transfer-RealPlay effectively maps control signals from virtual to real-world scenarios; and (2) entity transfer-although training labels originate solely from a car racing game, RealPlay generalizes to control diverse real-world entities, including bicycles and pedestrians, beyond vehicles. Project page can be found: https://wenqsun.github.io/RealPlay/

cs.CV↗

arXiv2025

UniFork: Exploring Modality Alignment for Unified Multimodal Understanding and Generation

Unified image understanding and generation has emerged as a promising paradigm in multimodal artificial intelligence. Despite recent progress, the optimal architectural design for such unified models remains an open challenge. In this work, we start by analyzing the modality alignment behaviors of task-specific expert models for understanding and generation, as well as current unified models. Our analysis reveals a crucial observation: understanding tasks benefit from a progressively increasing modality alignment across network depth, which helps build up semantic information for better comprehension; In contrast, generation tasks follow a different trend: modality alignment increases in the early layers but decreases in the deep layers to recover spatial details. These divergent alignment patterns create a fundamental conflict in fully shared Transformer backbones, where a uniform representational flow often leads to performance compromises across two tasks. Motivated by this finding, we introduce UniFork, a novel Y-shaped architecture that shares the shallow layers for cross-task representation learning, while employing task-specific branches in deeper layers to avoid task interference. This design effectively balances shared learning and task specialization. Through extensive ablation experiments, we demonstrate that Unifork consistently outperforms conventional fully shared Transformer architectures, and achieves performance on par with or better than task-specific models.

cs.CV↗