arXiv subjects

Junjie Li

Publications and source records attributed to Junjie Li.

At least 73 records · Page 4Linked to original sources

arXiv2026

Multipath Routing for Multi-Hop UAV Networks

Multi-hop uncrewed aerial vehicle (UAV) networks are promising to extend the terrestrial network coverage. Existing multi-hop UAV networks employ a single routing path by selecting the next-hop forwarding node in a hop-by-hop manner, which leads to local congestion and increases traffic delays. In this paper, a novel traffic-adaptive multipath routing method is proposed for multi-hop UAV networks, which enables each UAV to dynamically split and forward traffic flows across multiple next-hop neighbors, thus meeting latency requirements of diverse traffic flows in dynamic mobile environments. An on-time packet delivery ratio maximization problem is formulated to determine the traffic splitting ratios at each hop. This sequential decision-making problem is modeled as a decentralized partially observable Markov decision process (Dec-POMDP). To solve this Dec-POMDP, a novel multi-agent deep reinforcement leaning (MADRL) algorithm, termed Independent Proximal Policy Optimization with Dirichlet Modeling (IPPO-DM), is developed. Specifically, the IPPO serves as the core optimization framework, where the Dirichlet distribution is leveraged to parameterize a continuous stochastic policy network on the probability simplex, inherently ensuring feasible traffic splitting ratios. Simulation results demonstrate that IPPO-DM outperforms benchmark schemes in terms of both delivery latency guarantee and packet loss performance.

cs.MA↗

arXiv2026

Uncovering Intrinsic Capabilities: A Paradigm for Data Curation in Vision-Language Models

Large vision-language models (VLMs) achieve strong benchmark performance, but controlling their behavior through instruction tuning remains difficult. Reducing the budget of instruction tuning dataset often causes regressions, as heuristic strategies treat models as black boxes and overlook the latent capabilities that govern learning. We introduce Capability-Attributed Data Curation (CADC), a framework that shifts curation from task-specific heuristics to intrinsic capability analysis. CADC discovers intrinsic capabilities in an unsupervised manner from gradient-based learning trajectories, attributes training data to these capabilities via influence estimation, and curates capability-aware curricula through balanced selection and staged sequencing. This transforms black-box instruction tuning into a controllable, capability-driven process. With as little as 5% of the original data, CADC surpasses full-data training on multimodal benchmarks. These results validate intrinsic capabilities as the fundamental building blocks of model learning and establish CADC as a principle paradigm for instruction data curation.

cs.CV↗

arXiv2026

Learning Geometric Invariance for Gait Recognition

The goal of gait recognition is to extract identity-invariant features of an individual under various gait conditions, e.g., cross-view and cross-clothing. Most gait models strive to implicitly learn the common traits across different gait conditions in a data-driven manner to pull different gait conditions closer for recognition. However, relatively few studies have explicitly explored the inherent relations between different gait conditions. For this purpose, we attempt to establish connections among different gait conditions and propose a new perspective to achieve gait recognition: variations in different gait conditions can be approximately viewed as a combination of geometric transformations. In this case, all we need is to determine the types of geometric transformations and achieve geometric invariance, then identity invariance naturally follows. As an initial attempt, we explore three common geometric transformations (i.e., Reflect, Rotate, and Scale) and design a $\mathcal{R}$eflect-$\mathcal{R}$otate-$\mathcal{S}$cale invariance learning framework, named ${\mathcal{RRS}}$-Gait. Specifically, it first flexibly adjusts the convolution kernel based on the specific geometric transformations to achieve approximate feature equivariance. Then these three equivariant-aware features are respectively fed into a global pooling operation for final invariance-aware learning. Extensive experiments on four popular gait datasets (Gait3D, GREW, CCPG, SUSTech1K) show superior performance across various gait conditions.

cs.CV↗

arXiv2025

EM2LDL: A Multilingual Speech Corpus for Mixed Emotion Recognition through Label Distribution Learning

This study introduces EM2LDL, a novel multilingual speech corpus designed to advance mixed emotion recognition through label distribution learning. Addressing the limitations of predominantly monolingual and single-label emotion corpora \textcolor{black}{that restrict linguistic diversity, are unable to model mixed emotions, and lack ecological validity}, EM2LDL comprises expressive utterances in English, Mandarin, and Cantonese, capturing the intra-utterance code-switching prevalent in multilingual regions like Hong Kong and Macao. The corpus integrates spontaneous emotional expressions from online platforms, annotated with fine-grained emotion distributions across 32 categories. Experimental baselines using self-supervised learning models demonstrate robust performance in speaker-independent gender-, age-, and personality-based evaluations, with HuBERT-large-EN achieving optimal results. By incorporating linguistic diversity and ecological validity, EM2LDL enables the exploration of complex emotional dynamics in multilingual settings. This work provides a versatile testbed for developing adaptive, empathetic systems for applications in affective computing, including mental health monitoring and cross-cultural communication. The dataset, annotations, and baseline codes are publicly available at https://github.com/xingfengli/EM2LDL.

cs.CL↗

arXiv2025

Secure-Instruct: An Automated Pipeline for Synthesizing Instruction-Tuning Datasets Using LLMs for Secure Code Generation

Although Large Language Models (LLMs) show promising solutions to automated code generation, they often produce insecure code that threatens software security. Current approaches (e.g., SafeCoder) to improve secure code generation are limited by small, imbalanced instruction-tuning datasets. In this work, we present Secure-Instruct, a novel pipeline that automatically synthesizes high-quality vulnerable and secure code examples and instruction-tunes LLMs to align task description and secure code generation abilities. We evaluate Secure-Instruct on four representative LLMs using two security-related benchmarks: our own CWEBench and the existing CWEval. CWEBench comprises 93 scenarios on 44 CWEs, all without overlap with Secure-Instruct's synthetic instruction-tuning dataset, while CWEval covers 31 CWEs with 119 manually verified security-critical tasks. We find that Secure-Instruct improves both security and functional correctness in code generation. On CWEBench, Secure-Instruct substantially improves secure code generation, giving a 28.5% increase on average in secure ratio over the pre-trained models and outperforms SafeCoder by 12.6%. On CWEval, Secure-Instruct achieves an increase of 157.3% for CodeLlama-7B and 46.4% for Mistral-7B in Func-Sec@1 over pretrained models, and significantly outperforms SafeCoder.

cs.SE↗

arXiv2025

UVLM: Benchmarking Video Language Model for Underwater World Understanding

Recently, the remarkable success of large language models (LLMs) has achieved a profound impact on the field of artificial intelligence. Numerous advanced works based on LLMs have been proposed and applied in various scenarios. Among them, video language models (VidLMs) are particularly widely used. However, existing works primarily focus on terrestrial scenarios, overlooking the highly demanding application needs of underwater observation. To overcome this gap, we introduce UVLM, an under water observation benchmark which is build through a collaborative approach combining human expertise and AI models. To ensure data quality, we have conducted in-depth considerations from multiple perspectives. First, to address the unique challenges of underwater environments, we selected videos that represent typical underwater challenges including light variations, water turbidity, and diverse viewing angles to construct the dataset. Second, to ensure data diversity, the dataset covers a wide range of frame rates, resolutions, 419 classes of marine animals, and various static plants and terrains. Next, for task diversity, we adopted a structured design where observation targets are categorized into two major classes: biological and environmental. Each category includes content observation and change/action observation, totaling 20 distinct task types. Finally, we designed several challenging evaluation metrics to enable quantitative comparison and analysis of different methods. Experiments on two representative VidLMs demonstrate that fine-tuning VidLMs on UVLM significantly improves underwater world understanding while also showing potential for slight improvements on existing in-air VidLM benchmarks, such as VideoMME and Perception text. The dataset and prompt engineering will be released publicly.

cs.CV↗

arXiv2025

Scalable data-driven modeling of microstructure evolution by learning local dependency and spatiotemporal translation invariance rules in phase field simulation

Phase-field (PF) simulation provides a powerful framework for predicting microstructural evolution but suffers from prohibitive computational costs that severely limit accessible spatiotemporal scales in practical applications. While data-driven methods have emerged as promising approaches for accelerating PF simulations, existing methods require extensive training data from numerous evolution trajectories, and their inherent black-box nature raises concerns about long-term prediction reliability. This work demonstrates, through examples of grain growth and spinodal decomposition, that a minimalist Convolutional Neural Network (CNN) trained with a remarkably small dataset even from a single small-scale simulation can achieve seamless scalability to larger systems and reliable long-term predictions far beyond the temporal range of the training data. The key insight of this work lies in revealing that the success of CNN-based models stems from the alignment between their inductive biases and the physical priors of phase-field simulations specifically, locality and spatiotemporal translation invariance. Through effective receptive field analysis, we verify that the model captures these essential properties during training. Therefore, from a reductionist perspective, the surrogate model essentially establishes a spatiotemporally invariant regression mapping between a grid point's local environment and its subsequent state. Further analysis of the model's feature space demonstrates that microstructural evolution effectively represents a continuous redistribution of a finite set of local environments. When the model has already encountered nearly all possible local environments in the early-stage training data, it can reliably generalize to much longer evolution timescales, regardless of the dramatic changes in global microstructural morphology.

cond-mat.mtrl-sci↗

arXiv2025

Bell-State Quantum Holography with Metasurfaces

Metasurfaces composed of subwavelength nanostructures enable simultaneous control of polarization and wavefront, greatly enhancing holographic information capacity. Building on this capability, we extend holography into the quantum domain by experimentally realizing Bell-state holograms-distinct holographic images encoded in polarization-entangled Bell states of photon pairs. A polarization-multiplexed dielectric metasurface generates spatial modes conditioned on both input and output polarizations, entangling the holographic pattern with the two-photon state. To characterize these quantum holograms, we further develop quantum hologram tomography, reconstructing the full density matrix of the holographic state pixel by pixel. The reconstructed density-matrix hologram reveals tailor-made holographic symbols attached to individual Bell states through the metasurface, with contrast built up among the different Bell components as theory shows. This framework unifies metasurface photonics with quantum-state reconstruction and provides a scalable route toward high-dimensional quantum communication, encryption and information processing based on holographically encoded quantum light.

physics.optics↗

arXiv2025

Quasi-bound flat bands in the continuum

Bound states in the continuum (BICs) are widely known spatially localized states experimentally implemented as quasi-BICs. Although they emerged as a promising solution for achieving high-quality resonances in photonic structures, quasi-BICs are confined to a very narrow range in k-space and are highly sensitive to disorder. Here, we introduce quasi-bound flat bands in the continuum (quasi-BFICs), a class of optical states where Bloch modes are found within a photonic flat band, leading to a quasi-BIC behaviour at every k-point above the light line. We analytically and numerically demonstrate the origin of quasi-BFICs from the disorder-induced band folding, mode localization and multiple topological charges in k-space, and identify the optimal strength of structural disorder to maximise their generation probability. Angle-resolved transmission and Q-factor measurements confirm the existence of quasi-BFICs, opening new avenues for designing devices with high quality factor and wide-angle response, presenting a counterintuitive strategy that leverages disorder to enhance optical performance.

physics.optics↗

arXiv2025

Dependency-Aware Task Offloading in Multi-UAV Assisted Collaborative Mobile Edge Computing

This paper proposes a novel multi-unmanned aerial vehicle (UAV) assisted collaborative mobile edge computing (MEC) framework, where the computing tasks of terminal devices (TDs) can be decomposed into serial or parallel sub-tasks and offloaded to collaborative UAVs. We first model the dependencies among all sub-tasks as a directed acyclic graph (DAG) and design a two-timescale frame structure to decouple the sub-task interdependencies for sub-task scheduling. Then, a joint sub-task offloading, computational resource allocation, and UAV trajectories optimization problem is formulated, which aims to minimize the system cost, i.e., the weighted sum of the task completion delay and the system energy consumption. To solve this non-convex mixed-integer nonlinear programming (MINLP) problem, a penalty dual decomposition and successive convex approximation (PDD-SCA) algorithm is developed. Particularly, the original MINLP problem is equivalently transferred into a continuous form relying on PDD theory. By decoupling the resulting problem into three nested subproblems, the SCA method is further combined to recast the non-convex components and obtain desirable solutions. Numerical results demonstrate that: 1) Compared to the benchmark algorithms, the proposed scheme can significantly reduce the system cost, and thus realize an improved trade-off between task latency and energy consumption; 2) The proposed algorithm can achieve an efficient workload balancing for distributed computation across multiple UAVs.

cs.CY↗

arXiv2025

GCRank: A Generative Contextual Comprehension Paradigm for Takeout Ranking Model

The ranking stage serves as the central optimization and allocation hub in advertising systems, governing economic value distribution through eCPM and orchestrating the user-centric blending of organic and advertising content. Prevailing ranking models often rely on fragmented modules and hand-crafted features, limiting their ability to interpret complex user intent. This challenge is further amplified in location-based services such as food delivery, where user decisions are shaped by dynamic spatial, temporal, and individual contexts. To address these limitations, we propose a novel generative framework that reframes ranking as a context comprehension task, modeling heterogeneous signals in a unified architecture. Our architecture consists of two core components: the Generative Contextual Encoder (GCE) and the Generative Contextual Fusion (GCF). The GCE comprises three specialized modules: a Personalized Context Enhancer (PCE) for user-specific modeling, a Collective Context Enhancer (CCE) for group-level patterns, and a Dynamic Context Enhancer (DCE) for real-time situational adaptation. The GCF module then seamlessly integrates these contextual representations through low-rank adaptation. Extensive experiments confirm that our method achieves significant gains in critical business metrics, including click-through rate and platform revenue. We have successfully deployed our method on a large-scale food delivery advertising platform, demonstrating its substantial practical impact. This work pioneers a new perspective on generative recommendation and highlights its practical potential in industrial advertising systems.

cs.IR↗

arXiv2025

Reinforcement Learning with Verifiable Rewards Implicitly Incentivizes Correct Reasoning in Base LLMs

Recent advancements in long chain-of-thought (CoT) reasoning, particularly through the Group Relative Policy Optimization algorithm used by DeepSeek-R1, have led to significant interest in the potential of Reinforcement Learning with Verifiable Rewards (RLVR) for Large Language Models (LLMs). While RLVR promises to improve reasoning by allowing models to learn from free exploration, there remains debate over whether it truly enhances reasoning abilities or simply boosts sampling efficiency. This paper systematically investigates the impact of RLVR on LLM reasoning. We revisit Pass@K experiments and demonstrate that RLVR can extend the reasoning boundary for both mathematical and coding tasks. This is supported by our introduction of a novel evaluation metric, CoT-Pass@K, which captures reasoning success by accounting for both the final answer and intermediate reasoning steps. Furthermore, we present a theoretical framework explaining RLVR's incentive mechanism, demonstrating how it can encourage correct reasoning even when rewards are based solely on answer correctness. Our analysis of RLVR's training dynamics reveals that it incentivizes correct reasoning early in the process, with substantial improvements in reasoning quality confirmed through extensive evaluations. These findings provide strong evidence of RLVR's potential to enhance LLM reasoning, offering valuable insights into its mechanisms and performance improvements.

cs.AI↗

arXiv2025

Transient Stability Analysis of a Hybrid Grid-Forming and Grid-Following RES System Considering Multi-Mode Control Switching

The inherent control switching of renewable energy sources (RESs) during intricate transient processes introduces complexity to the dynamic behavior of modern power systems. This paper reveals the dynamic coupling between grid-forming (GFM)/grid-following (GFL)-based RES and dominant instability modes of the hybrid system. First, six control combinations are systematically investigated by pairing the two GFM-RES modes, normal control (NC) and current saturation (CS), with the three GFL-RES modes: normal control, low voltage ride-through (LVRT), and high voltage ride-through (HVRT). Based on switching system theory, the coupled power flow and dynamic motion models are developed considering multi-mode switching characteristics. It is revealed that the hybrid system exhibits two distinct instability modes when the GFM-RES and GFL-RES exceed their P-f and V-f desynchronization boundaries, respectively. The two-dimensional spatiotemporal damping characteristics of GFL-RES induced by GFM-RES are also uncovered for the first time. A novel criterion is proposed to quantify the impact of GFM-RES on GFL-RES dynamics, capturing both its stabilizing and destabilizing effects under different control combinations. High-fidelity electromagnetic transient simulations validate the correctness of the analysis framework.

eess.SY↗

arXiv2025

An Exploratory Study on Fine-Tuning Large Language Models for Secure Code Generation

AI-powered coding assistants such as GitHub's Copilot and OpenAI's ChatGPT have achieved notable success in automating code generation. However, these tools rely on pre-trained Large Language Models (LLMs) that are typically trained on human-written code sourced from open-source project hosting sites like GitHub, which often contains inherent security vulnerabilities. These vulnerabilities may then be mirrored in the code generated by these LLMs, a critical risk revealed and highlighted by recent empirical studies. In this work, we present an exploratory study on whether fine-tuning pre-trained LLMs on datasets of vulnerability-fixing commits can promote secure code generation. We explored full fine-tuning and two parameter-efficient fine-tuning techniques (LoRA and IA3) on four pre-trained LLMs for code generation. We crawled a fine-tuning dataset (14,622 C/C++ files) for secure code generation by collecting code fixes of confirmed vulnerabilities from open-source repositories. Our evaluation dataset comprises 52 vulnerability scenarios designed to cover the top most dangerous C/C++ CWEs. Our exploration reveals that fine-tuning LLMs using PEFT techniques can enhance secure code generation. We observe maximum improvements in security of 6.4% in C language and 5.0% in C++ language. In addition, we compared between the fine-tuning approaches and the prompt-based approaches. The LoRA-tuned models outperform the prompt-based approaches in secure code generation. We found that fine-tuning with function-level and block-level datasets achieves the best secure code generation performance, compared to the alternatives (file-level and line-level).

cs.SE↗

arXiv2025

SMT(LIA) Sampling with High Diversity

Satisfiability Modulo Linear Integer Arithmetic, SMT(LIA) for short, is pivotal across various critical domains. Previous research has primarily focused on SMT solving techniques. However, in practical applications such as software and hardware testing, there is a need to generate a diverse set of solutions for use as test inputs. We have developed the first sampling framework that integrates local search with CDCL(T) techniques, named HighDiv, capable of generating a highly diverse set of solutions for constraints under linear integer theory. Initially, in the local search phase, we introduced a novel operator called boundary-aware movement. This operator performs random moves by considering the current state's constraints on variables, thereby enhancing the diversity of variables during the search process. Furthermore, we have conducted an in-depth study of the preprocessing and variable initialization mechanisms within the framework, which significantly enhances the efficiency of subsequent local searches. Lastly, we use the solutions obtained from local search sampling as additional constraints to further explore the solution space using the stochastic CDCL(T) method. Experimental results demonstrate that \HighDiv generates solutions with greater diversity compared to the state-of-the-art SMT(LIA) sampling tool, MeGASampler.

cs.LO↗

arXiv2025

FireRedChat: A Pluggable, Full-Duplex Voice Interaction System with Cascaded and Semi-Cascaded Implementations

Full-duplex voice interaction allows users and agents to speak simultaneously with controllable barge-in, enabling lifelike assistants and customer service. Existing solutions are either end-to-end, difficult to design and hard to control, or modular pipelines governed by turn-taking controllers that ease upgrades and per-module optimization; however, prior modular frameworks depend on non-open components and external providers, limiting holistic optimization. In this work, we present a complete, practical full-duplex voice interaction system comprising a turn-taking controller, an interaction module, and a dialogue manager. The controller integrates streaming personalized VAD (pVAD) to suppress false barge-ins from noise and non-primary speakers, precisely timestamp primary-speaker segments, and explicitly enable primary-speaker barge-ins; a semantic end-of-turn detector improves stop decisions. It upgrades heterogeneous half-duplex pipelines, cascaded, semi-cascaded, and speech-to-speech, to full duplex. Using internal models, we implement cascaded and semi-cascaded variants; the semi-cascaded one captures emotional and paralinguistic cues, yields more coherent responses, lowers latency and error propagation, and improves robustness. A dialogue manager extends capabilities via tool invocation and context management. We also propose three system-level metrics, barge-in, end-of-turn detection accuracy, and end-to-end latency, to assess naturalness, control accuracy, and efficiency. Experiments show fewer false interruptions, more accurate semantic ends, and lower latency approaching industrial systems, enabling robust, natural, real-time full-duplex interaction. Demos: https://fireredteam.github.io/demos/firered_chat.

cs.SD↗

arXiv2025

FireRedTTS-2: Towards Long Conversational Speech Generation for Podcast and Chatbot

Current dialogue generation approaches typically require the complete dialogue text before synthesis and produce a single, inseparable speech containing all voices, making them unsuitable for interactive chat; moreover, they suffer from unstable synthesis, inaccurate speaker transitions, and incoherent prosody. In this work, we present FireRedTTS-2, a long-form streaming TTS system for multi-speaker dialogue generation, delivering stable, natural speech with reliable speaker switching and context-aware prosody. A new 12.5Hz streaming speech tokenizer accelerates training and inference, extends maximum dialogue length, encodes richer semantics to stabilize text-to-token modeling and supports high-fidelity streaming generation for real-time applications. We adopt a text-speech interleaved format, concatenating speaker-labeled text with aligned speech tokens in chronological order, and model it with a dual-transformer: a large decoder-only transformer predicts tokens at the first layer, and a smaller one completes subsequent layers. Experimental results show that FireRedTTS-2 integrates seamlessly with chat frameworks and, with minimal fine-tuning, produces emotionally expressive speech guided by implicit contextual cues. In podcast generation, it surpasses existing systems including MoonCast, Zipvoice-Dialogue, and MOSS-TTSD in objective intelligibility, speaker-turn reliability, and perceived naturalness with context-consistent prosody. Our demos are available at https://fireredteam.github.io/demos/firered_tts_2.

cs.SD↗

arXiv2025

Improved PLL Design for Transient Stability Enhancement of Grid Following Converters Based on Lyapunov Method

Fluctuations in phase angle and frequency under large disturbances can lead to loss of synchronism (LOS) in grid-following (GFL) converters. The power angle and frequency of synchronous generators (SGs) correspond to rotor position and speed, whereas those of converters lack a direct physical counterpart in the real world and can thus be directly adjusted by control methods to prevent loss of synchronization. In this paper, an improved phase-locked loop (PLL) design with reset control for GFL converters is proposed to enhance transient stability. The stability domain (SD) of a GFL converter is first analyzed, and three forms of SD are identified under different short circuit ratios. Secondly, based on the characteristics of the three SD forms, two PLL-reset methods are proposed, including omega reset and omega&delta reset. Thirdly, to provide the triggering conditions for the PLL-reset control, the Lyapunov function of the GFL converter is constructed based on three methods: the approximation-based Lyapunov method, the Zubov method, and the analytical trajectory reversing method. All these methods are immune to the negative damping problem of PLL dynamics, which makes traditional energy-perspective Lyapunov functions invalid. Finally, the estimation accuracy of the three Lyapunov-based methods is analyzed, and the effectiveness of the PLL-reset control is verified in single-machine and multi-machine case studies.

eess.SY↗