Institution profile

Great Bay University

Academic institutionasia · cn
Official website
Research library274linked papers
Opportunities0open roles
Selected work

Representative Papers

A Comprehensive Survey on Vector Database: Storage and Retrieval Technique, Challenge

Oct 18, 2023arXiv.org

Managing and retrieving high-dimensional vector data poses significant challenges, particularly as traditional databases fail to meet performance requirements and the need for tight integration with large language models (LLMs) intensifies. Method: This paper systematically surveys four major approximate nearest neighbor search (ANNS) paradigms—hashing, tree-based indexing, graph-based methods (e.g., HNSW), and quantization (PQ/SQ)—and integrates hybrid optimization strategies. Contribution/Results: It introduces, for the first time, a “Four-Dimensional Methodology” framework tailored for industrial deployment of vector databases, analyzing trade-offs among accuracy, latency, memory footprint, and scalability. The work constructs a structured knowledge graph covering 200+ ANNS algorithms and proposes a novel paradigm for deep synergy between vector databases and LLMs. Collectively, these contributions provide both theoretical foundations and practical guidelines for system selection, architectural design, and development of AI-native database systems.

62 citations3 influentialRead paper

Benchmarking Cross-Domain Audio-Visual Deception Detection

May 11, 2024arXiv.org

Current audio-visual spoofing detection methods suffer from poor cross-scenario generalization and lack a standardized cross-domain evaluation benchmark. To address this, we introduce the first unified, standardized cross-domain benchmark for audio-visual spoofing detection, supporting both single-source-to-single-target and multi-source-to-single-target domain adaptation settings. We propose MM-IDGM, a gradient-coordinated optimization algorithm, and Attention-Mixer, a novel multimodal fusion architecture. Additionally, we design three novel multi-source domain sampling strategies and integrate OpenSMILE/ResNet-50 feature extractors with CNN/RNN/Transformer backbones. Extensive experiments demonstrate that our approach achieves an average accuracy improvement of 5.2% under the multi-source-to-single-target setting, significantly enhancing cross-domain generalization. The benchmark and methodology provide a reproducible, comparable, and realistic evaluation framework for practical deployment.

2 citationsRead paper

MacWilliams Theory over Zk and nu-functions over Lattices

Apr 24, 2025

This paper addresses two central problems: (1) generalizing the MacWilliams identity to $m$-fold codes over $mathbb{Z}_k$, and verifying whether its complete weight enumerator form extends to finitely generated rings $mathbb{Z}_k[xi]$; and (2) assessing the general validity of Solé’s (1995) MacWilliams-type conjecture concerning the $ u$-function on lattices. Employing algebraic coding theory, lattice theory, character-theoretic methods, and modular form analysis, we derive the first explicit formula for the $ u$-function associated with ternary codes’ lattices. We rigorously prove that Solé’s conjecture holds only for binary code lattices and construct multiple counterexamples demonstrating its failure in general. Our results establish a unified framework for generalized MacWilliams identities over $mathbb{Z}_k$ and $mathbb{Z}_k[xi]$, and fully characterize the fundamental dichotomy—binary versus ternary—in the lattice-theoretic analogues of the $ u$-function.

1 citations1 influentialRead paper

UniBiDex: A Unified Teleoperation Framework for Robotic Bimanual Dexterous Manipulation

Jan 08, 2026arXiv.org

This work proposes the first unified dual-arm teleoperation framework that integrates both virtual reality (VR) and master-slave modalities to address key challenges in heterogeneous input device integration, real-time performance, safety, and coordinated motion. The framework employs a shared control stack to ensure consistent kinematic processing and enforce safety constraints, complemented by a null-space optimization strategy that guarantees collision-free and singularity-avoiding cooperative motion. It enables real-time, contact-rich dexterous manipulation and is accompanied by open-sourced hardware and software to facilitate the collection of high-quality demonstration data. In a long-horizon kitchen organization task comprising five subtasks, the proposed method significantly outperforms strong baselines in task success rate, trajectory smoothness, and robustness.

1 citationsRead paper

Vulnerabilities in AI-generated Image Detection: The Challenge of Adversarial Attacks

Jul 30, 2024arXiv.org

This work addresses the insufficient robustness of AI-generated image (AIGI) detectors against adversarial attacks, presenting the first systematic evaluation of their vulnerability under both white-box and black-box settings. We propose the Frequency-domain–Bayesian Attack (FPBA) framework: it generates highly transferable adversarial perturbations in the frequency domain and incorporates a post-training Bayesian surrogate model to approximate the target detector’s uncertainty distribution—enabling efficient black-box attacks across architectures (CNNs and ViTs), generative models, and defensive mechanisms. Experiments demonstrate that FPBA significantly degrades detection accuracy across diverse AIGI detectors, multiple generative models (e.g., Stable Diffusion, DALL·E), and state-of-the-art defenses (e.g., JPEG compression, feature squeezing). Crucially, FPBA provides the first empirical evidence of cross-generator evasion capability. Our work establishes a new benchmark and methodological foundation for advancing the robustness evaluation and defense of AIGI detection systems.

1 citationsRead paper
Recent publications

Latest Papers