arXiv subjects

Yanming Xiu

Publications and source records attributed to Yanming Xiu.

12 recordsLinked to original sources

arXiv2026

Real-Time Visual Obstruction Detection in Surgical Augmented Reality

Surgical augmented reality (AR) can provide contextual guidance by overlaying virtual annotations, tool cues, and procedural information onto the surgical workspace. However, the virtual content may obstruct task-relevant real-world information, such as surgical instruments, and interfere with users' perception during time-sensitive surgical tasks. In this paper, we investigate visual obstruction detection for surgical AR and present a latency-aware pipeline that combines vision-language model (VLM)-based surgical-object recognition with segmentation-based obstruction reasoning. To reduce inference overhead, the system adopts a cascaded small-to-large VLM architecture with segmentation-guided early exiting and attention-based visual token pruning. The small VLM handles easy frames when its key-object prediction is supported by segmentation consistency, while difficult frames are forwarded to a large VLM with pruned visual tokens. We construct a pseudo-AR surgical obstruction detection benchmark by overlaying virtual content onto surgical-tool images and labeling whether the virtual content obstructs task-relevant instruments. Evaluation results show that the proposed system achieves 87.43% obstruction detection accuracy with an average end-to-end latency of 479 ms, reducing latency by 62.90% compared with a cloud large-model baseline. These results demonstrate the feasibility of latency-aware obstruction detection for surgical AR and motivate future work on dynamic surgical videos, multi-object scenes, and clinically grounded AR guidance content.

cs.CV↗

arXiv2026

Is It Real? Exploiting Virtual-Physical Discrimination Vulnerability in Mixed Reality

Consumer mixed reality (MR) headsets seamlessly blend virtual content into physical environments with sufficient fidelity that users may be unable to distinguish virtual objects from physical ones. We identify this virtual-physical discrimination vulnerability as an exploitable security primitive. Through speculative design workshops with 12 experts from cybersecurity and MR/HCI, we develop a taxonomy of virtual-physical confusion attacks and implement four proof-of-concept attacks on Apple Vision Pro, evaluating them with 26 participants in realistic MR tasks. All four attacks altered user behavior, with success rates ranging from 85% to 100%, producing misdirected interactions, misjudged object identities, biased purchasing decisions, and altered navigation paths. Notably, the most successful attacks were also the hardest to detect according to participants' subjective ratings. Even participants who recognized virtual content still complied behaviorally, and no participant attributed anomalous events to adversarial causes. We propose platform-level provenance, interaction gating, and user education as countermeasures.

cs.HC↗

arXiv2026

Benchmarking Vision-Language Models under Contradictory Virtual Content Attacks in Augmented Reality

Augmented reality (AR) has rapidly expanded over the past decade. As AR becomes increasingly integrated into daily life, its security and reliability emerge as critical challenges. Among various threats, contradictory virtual content attacks, where malicious or inconsistent virtual elements are introduced into the user's view, pose a unique risk by misleading users, creating semantic confusion, or delivering harmful information. In this work, we systematically model such attacks and present ContrAR, a novel benchmark for evaluating the robustness of vision-language models (VLMs) against virtual content manipulation and contradiction in AR. ContrAR contains 312 real-world AR videos validated by 10 human participants. We further benchmark 11 VLMs, including both commercial and open-source models. Experimental results reveal that while current VLMs exhibit reasonable understanding of contradictory virtual content, room still remains for improvement in detecting and reasoning about adversarial content manipulations in AR environments. Moreover, balancing detection accuracy and latency remains challenging.

cs.CV↗

arXiv2026

User Prompting Strategies and Prompt Enhancement Methods for Open-Set Object Detection in XR Environments

Open-set object detection (OSOD) localizes objects while identifying and rejecting unknown classes at inference. While recent OSOD models perform well on benchmarks, their behavior under realistic user prompting remains underexplored. In interactive XR settings, user-generated prompts are often ambiguous, underspecified, or overly detailed. To study prompt-conditioned robustness, we evaluate two OSOD models, GroundingDINO and YOLO-E, on real-world XR images and simulate diverse user prompting behaviors using vision-language models. We consider four prompt types: standard, underdetailed, overdetailed, and pragmatically ambiguous, and examine the impact of two enhancement strategies on these prompts. Results show that both models exhibit stable performance under underdetailed and standard prompts, while they suffer degradation under ambiguous prompts. Overdetailed prompts primarily affect GroundingDINO. Prompt enhancement substantially improves robustness under ambiguity, yielding gains exceeding 55% mIoU and 41% average confidence. Based on the findings, we propose several prompting strategies and prompt enhancement methods for OSOD models in XR environments.

cs.CV↗

arXiv2026

Can a Unimodal Language Agent Provide Preferences to Tune a Multimodal Vision-Language Model?

To explore a more scalable path for adding multimodal capabilities to existing LLMs, this paper addresses a fundamental question: Can a unimodal LLM, relying solely on text, reason about its own informational needs and provide effective feedback to optimize a multimodal model? To answer this, we propose a method that enables a language agent to give feedback to a vision-language model (VLM) to adapt text generation to the agent's preferences. Our results from different experiments affirm this hypothesis, showing that LLM preference feedback significantly enhances VLM descriptions. Using our proposed method, we find that the VLM can generate multimodal scene descriptions to help the LLM better understand multimodal context, leading to improvements of maximum 13% in absolute accuracy compared to the baseline multimodal approach. Furthermore, a human study validated our AI-driven feedback, showing a 64.6% preference alignment rate between the LLM's choices and human judgments. Extensive experiments provide insights on how and why the method works and its limitations.

cs.CL↗

arXiv2025

Demonstrating Visual Information Manipulation Attacks in Augmented Reality: A Hands-On Miniature City-Based Setup

Augmented reality (AR) enhances user interaction with the real world but also presents vulnerabilities, particularly through Visual Information Manipulation (VIM) attacks. These attacks alter important real-world visual cues, leading to user confusion and misdirected actions. In this demo, we present a hands-on experience using a miniature city setup, where users interact with manipulated AR content via the Meta Quest 3. The demo highlights the impact of VIM attacks on user decision-making and underscores the need for effective security measures in AR systems. Future work includes a user study and cross-platform testing.

cs.HC↗

arXiv2025

Say It, See It: A Systematic Evaluation on Speech-Based 3D Content Generation Methods in Augmented Reality

As augmented reality (AR) applications increasingly require 3D content, generative pipelines driven by natural input such as speech offer an alternative to manual asset creation. In this work, we design a modular, edge-assisted architecture that supports both direct text-to-3D and text-image-to-3D pathways, enabling interchangeable integration of state-of-the-art components and systematic comparison of their performance in AR settings. Using this architecture, we implement and evaluate four representative pipelines through an IRB-approved user study with 11 participants, assessing six perceptual and usability metrics across three object prompts. Overall, text-image-to-3D pipelines deliver higher generation quality: the best-performing pipeline, which used FLUX for image generation and Trellis for 3D generation, achieved an average satisfaction score of 4.55 out of 5 and an intent alignment score of 4.82 out of 5. In contrast, direct text-to-3D pipelines excel in speed, with the fastest, Shap-E, completing generation in about 20 seconds. Our results suggest that perceptual quality has a greater impact on user satisfaction than latency, with users tolerating longer generation times when output quality aligns with expectations. We complement subjective ratings with system-level metrics and visual analysis, providing practical insights into the trade-offs of current 3D generation methods for real-world AR deployment.

cs.HC↗

arXiv2025

A Neurosymbolic Framework for Interpretable Cognitive Attack Detection in Augmented Reality

Augmented Reality (AR) enriches human perception by overlaying virtual elements onto the physical world. However, this tight coupling between virtual and real content makes AR vulnerable to cognitive attacks: manipulations that distort users' semantic understanding of the environment. Existing detection methods largely focus on visual inconsistencies at the pixel or image level, offering limited semantic reasoning or interpretability. To address these limitations, we introduce CADAR, a neuro-symbolic framework for cognitive attack detection in AR that integrates neural and symbolic reasoning. CADAR fuses multimodal vision-language representations from pre-trained models into a perception graph that captures objects, relations, and temporal contextual salience. Building on this structure, a particle-filter-based statistical reasoning module infers anomalies in semantic dynamics to reveal cognitive attacks. This combination provides both the adaptability of modern vision-language models and the interpretability of probabilistic symbolic reasoning. Preliminary experiments on an AR cognitive-attack dataset demonstrate consistent advantages over existing approaches, highlighting the potential of neuro-symbolic methods for robust and interpretable AR security.

cs.CV↗

arXiv2025

Toward Safe, Trustworthy and Realistic Augmented Reality User Experience

As augmented reality (AR) becomes increasingly integrated into everyday life, ensuring the safety and trustworthiness of its virtual content is critical. Our research addresses the risks of task-detrimental AR content, particularly that which obstructs critical information or subtly manipulates user perception. We developed two systems, ViDDAR and VIM-Sense, to detect such attacks using vision-language models (VLMs) and multimodal reasoning modules. Building on this foundation, we propose three future directions: automated, perceptually aligned quality assessment of virtual content; detection of multimodal attacks; and adaptation of VLMs for efficient and user-centered deployment on AR devices. Overall, our work aims to establish a scalable, human-aligned framework for safeguarding AR experiences and seeks feedback on perceptual modeling, multimodal AR content implementation, and lightweight model adaptation.

cs.CV↗

arXiv2025

Detecting Visual Information Manipulation Attacks in Augmented Reality: A Multimodal Semantic Reasoning Approach

The virtual content in augmented reality (AR) can introduce misleading or harmful information, leading to semantic misunderstandings or user errors. In this work, we focus on visual information manipulation (VIM) attacks in AR, where virtual content changes the meaning of real-world scenes in subtle but impactful ways. We introduce a taxonomy that categorizes these attacks into three formats: character, phrase, and pattern manipulation, and three purposes: information replacement, information obfuscation, and extra wrong information. Based on the taxonomy, we construct a dataset, AR-VIM, which consists of 452 raw-AR video pairs spanning 202 different scenes, each simulating a real-world AR scenario. To detect the attacks in the dataset, we propose a multimodal semantic reasoning framework, VIM-Sense. It combines the language and visual understanding capabilities of vision-language models (VLMs) with optical character recognition (OCR)-based textual analysis. VIM-Sense achieves an attack detection accuracy of 88.94% on AR-VIM, consistently outperforming vision-only and text-only baselines. The system achieves an average attack detection latency of 7.07 seconds in a simulated video processing framework and 7.17 seconds in a real-world evaluation conducted on a mobile Android AR application.

cs.CV↗

arXiv2025

ViDDAR: Vision Language Model-Based Task-Detrimental Content Detection for Augmented Reality

In Augmented Reality (AR), virtual content enhances user experience by providing additional information. However, improperly positioned or designed virtual content can be detrimental to task performance, as it can impair users' ability to accurately interpret real-world information. In this paper we examine two types of task-detrimental virtual content: obstruction attacks, in which virtual content prevents users from seeing real-world objects, and information manipulation attacks, in which virtual content interferes with users' ability to accurately interpret real-world information. We provide a mathematical framework to characterize these attacks and create a custom open-source dataset for attack evaluation. To address these attacks, we introduce ViDDAR (Vision language model-based Task-Detrimental content Detector for Augmented Reality), a comprehensive full-reference system that leverages Vision Language Models (VLMs) and advanced deep learning techniques to monitor and evaluate virtual content in AR environments, employing a user-edge-cloud architecture to balance performance with low latency. To the best of our knowledge, ViDDAR is the first system to employ VLMs for detecting task-detrimental content in AR settings. Our evaluation results demonstrate that ViDDAR effectively understands complex scenes and detects task-detrimental content, achieving up to 92.15% obstruction detection accuracy with a detection latency of 533 ms, and an 82.46% information manipulation content detection accuracy with a latency of 9.62 s.

cs.CV↗

arXiv2025

Advancing the Understanding and Evaluation of AR-Generated Scenes: When Vision-Language Models Shine and Stumble

Augmented Reality (AR) enhances the real world by integrating virtual content, yet ensuring the quality, usability, and safety of AR experiences presents significant challenges. Could Vision-Language Models (VLMs) offer a solution for the automated evaluation of AR-generated scenes? Could Vision-Language Models (VLMs) offer a solution for the automated evaluation of AR-generated scenes? In this study, we evaluate the capabilities of three state-of-the-art commercial VLMs -- GPT, Gemini, and Claude -- in identifying and describing AR scenes. For this purpose, we use DiverseAR, the first AR dataset specifically designed to assess VLMs' ability to analyze virtual content across a wide range of AR scene complexities. Our findings demonstrate that VLMs are generally capable of perceiving and describing AR scenes, achieving a True Positive Rate (TPR) of up to 93% for perception and 71% for description. While they excel at identifying obvious virtual objects, such as a glowing apple, they struggle when faced with seamlessly integrated content, such as a virtual pot with realistic shadows. Our results highlight both the strengths and the limitations of VLMs in understanding AR scenarios. We identify key factors affecting VLM performance, including virtual content placement, rendering quality, and physical plausibility. This study underscores the potential of VLMs as tools for evaluating the quality of AR experiences.

cs.CV↗