Institution profile

Singapore Management University

Academic institutionasia · sg
Official website
Research library986linked papers
Opportunities0open roles
Selected work

Representative Papers

AnomalyCLIP: Object-agnostic Prompt Learning for Zero-shot Anomaly Detection

Oct 29, 2023International Conference on Learning Representations

Zero-shot anomaly detection (ZSAD) aims to identify anomalies across domains without access to target-domain training samples; however, its generalizability is severely hindered by substantial discrepancies in foreground objects, anomaly appearances, and background distributions. To address this, we propose a CLIP-based universal ZSAD framework. Our method introduces the first object-agnostic text prompt learning mechanism, decoupling foreground semantics from normal/abnormal discrimination modeling. We further design learnable, domain-agnostic “normal” and “abnormal” text prompts and leverage vision-language feature alignment to enable zero-shot anomaly scoring and pixel-level localization. Critically, our approach requires no target-domain annotations or fine-tuning. Extensive experiments across 17 industrial defect and medical imaging datasets demonstrate significant improvements over existing ZSAD methods, achieving both strong cross-domain generalization and high localization accuracy.

114 citations20 influentialRead paper

An Empirical Study of Automated Vulnerability Localization with Large Language Models

Mar 30, 2024arXiv.org

This work systematically evaluates the effectiveness of large language models (LLMs) for line-level vulnerability localization (AVL)—a task lacking comprehensive empirical investigation. Experiments are conducted on BigVul (C/C++) and smart contract vulnerability datasets, covering over ten code-understanding LLMs (60M–16B parameters) spanning encoder-only, encoder-decoder, and decoder-only architectures, under zero-shot, one-shot, discriminative fine-tuning, and generative fine-tuning paradigms. Key contributions include: (1) the first empirical demonstration that discriminative fine-tuning substantially outperforms existing approaches; (2) the proposal of sliding-window context partitioning and right-forward embedding to mitigate context-length limitations; and (3) strong cross-CWE and cross-project generalization, yielding significant improvements in localization accuracy and surpassing state-of-the-art methods.

25 citations2 influentialRead paper

Examining Augmented Virtuality Impairment Simulation for Mobile App Accessibility Design

May 02, 2019International Conference on Human Factors in Computing Systems

This study addresses the limitations of existing mobile application accessibility evaluation methods for cataract users—namely, perceptual distortion and insufficient empathic engagement. To bridge this gap, we propose an Augmented Virtuality (AV)-driven design support methodology. We introduce Empath-D, the first system to employ AV for real-time, interactive simulation of cataract-induced visual impairment. Empath-D integrates eye-tracking and gesture interaction simulation, on-device real-time rendering, and an accessibility-oriented human factors evaluation framework, enabling embodied understanding of authentic usage challenges. Compared to conventional guideline-based tools, Empath-D increases defect detection count by 37% and improves identification accuracy by 29%. User interviews confirm its effectiveness in significantly enhancing designers’ empathic capacity and grounding design decisions in lived experience. This work establishes a novel application paradigm for AV in accessible human–computer interaction.

14 citationsRead paper

Empath-D: VR-based Empathetic App Design for Accessibility

Jun 10, 2018ACM SIGMOBILE International Conference on Mobile Systems, Applications, and Services

Addressing the challenge of evaluating accessibility in design, this paper introduces a VR–smartphone collaborative empathic evaluation system that enables designers to simulate, in real time and in situ, the interactive experiences of users with visual, auditory, and motor impairments when using real-world mobile applications. The system pioneers lightweight I/O stream perturbation and hand-motion visualization techniques, achieving high-fidelity impairment modeling—including visual blurring, audio attenuation, and touch jitter—while preserving interaction fluency. Leveraging VR-based rendering, low-latency mobile communication, and precise hand gesture tracking, it attains an average interaction accuracy of 92.3% and subjectively realistic user experiences. Experimental results demonstrate that its impairment simulation fidelity matches that of dedicated hardware solutions. This work establishes a novel, efficient, low-cost, and scalable paradigm for inclusive design evaluation.

10 citationsRead paper

Test smells in LLM-Generated Unit Tests

Oct 14, 2024arXiv.org

This study addresses the understudied issue of test smells in large language model (LLM)-generated unit tests. Method: We conduct the first large-scale empirical analysis of test smells across 20,505 Java class-level test suites drawn from five sources: human-written tests, EvoSuite-generated tests, and LLM-generated tests from GPT-3.5, GPT-4, Mistral, and Mixtral. Our multi-benchmark, cross-model analysis framework encompasses over 770,000 test cases, leveraging dual smell-detection tools—TsDetect and JNose—across 34,635 open-source projects and the TestBench benchmark. Contribution/Results: We identify prevalent smells—including Assertion Roulette and Magic Number Test—in LLM-generated tests; their occurrence patterns are significantly influenced by prompting strategies, context length, and model scale. Notably, LLM-generated tests exhibit smell profiles closer to human-written tests than to search-based software testing (SBST) outputs, suggesting potential training data contamination. These findings provide critical empirical grounding for developing smell-aware test generation frameworks.

4 citationsRead paper
Recent publications

Latest Papers