arXiv subjects

Hsiang-Ting Chen

Publications and source records attributed to Hsiang-Ting Chen.

At least 19 recordsLinked to original sources

arXiv2026

DisMix: Order-Aware Mixup for Medical Imaging via Disentangling Ordinal and Non-Ordinal Features

Image mixup is a widely adopted data augmentation strategy, yet it is ill-suited for ordinal classification tasks such as medical disease grading, where labels encode a progression of severity. By indiscriminately blending disease-severity cues (ordinal) with appearance-level variation (non-ordinal), standard mixup produces samples that distort the very ordinal structure that underpins clinical severity grading. We introduce DisMix, an order-aware mixup framework for ordinal classification. DisMix disentangles ordinal and non-ordinal features via a dual-codebook VQ-VAE, allowing each subspace to be mixed independently: ordinal codes are interpolated to produce meaningful intermediate ranks, while non-ordinal codes are varied to introduce appearance diversity without corrupting the ordinal signal. Across four medical imaging datasets, DisMix shows the best aggregate performance among six image mixup baselines paired with six ordinal classifiers and remains effective under data scarcity and clinical grading variability.

cs.CV↗

arXiv2026

DeltaMCP: Incremental Regeneration via Spec-Aware Transformation for MCP servers

The rapid development of LLMs coupled with the introduction of Model Context Protocol (MCP) has revolutionized how intelligent agents interact with APIs through deterministic and structured methods \cite{ModelContextProtocolIntro2025}. While some existing systems like AutoMCP attempt to automate a previously completely manual process of generating MCP servers, they fail to address the recurring challenge of maintaining synchronization between evolving enterprise-level APIs and their corresponding MCP toolset implementation \cite{mastouri2025makingrestapisagentready}. This paper introduces DeltaMCP, a specification-aware, incremental regeneration tool for enterprise-grade MCP servers. DeltaMCP enables developers to only update the affected tooling of MCP servers, given a new release of it's corresponding service's OpenAPI specification. Using Azure REST API specifications as the evaluation dataset, DeltaMCP is benchmarked against baseline full generation methods on generation quality and system performance. The results demonstrate the reduction in developer overhead through DeltaMCP whilst improving maintainability and version consistency. This research offers a scalable approach for enterprises seeking to maintain high-fidelity, up-to-date MCP server infrastructures for LLM-based systems.

cs.SE↗

arXiv2026

Understanding Annotation Error Propagation and Learning an Adaptive Policy for Expert Intervention in Barrett's Video Segmentation

Accurate annotation of endoscopic videos is essential yet time-consuming, particularly for challenging datasets such as dysplasia in Barrett's esophagus, where the affected regions are irregular and lack clear boundaries. Semi-automatic tools like Segment Anything Model 2 (SAM2) can ease this process by propagating annotations across frames, but small errors often accumulate and reduce accuracy, requiring expert review and correction. To address this, we systematically study how annotation errors propagate across different prompt types, namely masks, boxes, and points, and propose Learning-to-Re-Prompt (L2RP), a cost-aware framework that learns when and where to seek expert input. By tuning a human-cost parameter, our method balances annotation effort and segmentation accuracy. Experiments on a private Barrett's dysplasia dataset and the public SUN-SEG benchmark demonstrate improved temporal consistency and superior performance over baseline strategies.

cs.CV↗

arXiv2026

EndoExtract: Co-Designing Structured Text Extraction from Endometriosis Ultrasound Reports

Endometriosis ultrasound reports are often unstructured free-text documents that require manual abstraction for downstream tasks such as analytics, machine learning model training, and clinical auditing. We present \textbf{EndoExtract}, an on-premise LLM-powered system that extracts structured data from these reports and surfaces interpretive fields for human review. Through contextual inquiry with research assistants, we identified key workflow pain points: asymmetric trust between numerical and interpretive fields, repetitive manual highlighting, fatigue from sustained comparison, and terminology inconsistency across radiologists. These findings informed an interface that surfaces only interpretive fields for mandatory review, automatically highlights source evidence within PDFs, and separates batch extraction from human-paced verification. A formative workshop revealed that \textbf{EndoExtract} supports a shift from field-by-field data entry to supervisory validation, though participants noted risks of over-skimming and challenges in managing missing data.

cs.HC↗

arXiv2026

Deep Multimodal Learning with Missing Modality: A Survey

During multimodal model training and testing, certain data modalities may be absent due to sensor limitations, cost constraints, privacy concerns, or data loss, negatively affecting performance. Multimodal learning techniques designed to handle missing modalities can mitigate this by ensuring model robustness even when some modalities are unavailable. This survey reviews recent progress in Multimodal Learning with Missing Modality (MLMM), focusing on deep learning methods. It provides the first comprehensive survey that covers the motivation and distinctions between MLMM and standard multimodal learning setups, followed by a detailed analysis of current methods, applications, and datasets, concluding with challenges and future directions.

cs.CV↗

arXiv2026

Who Fails Where? LLM and Human Error Patterns in Endometriosis Ultrasound Report Extraction

In this study, we evaluate a locally-deployed large-language model (LLM) to convert unstructured endometriosis transvaginal ultrasound (eTVUS) scan reports into structured data for imaging informatics workflows. Across 49 eTVUS reports, we compared three LLMs (7B/8B and a 20B-parameter model) against expert human extraction. The 20B model achieved a mean accuracy of 86.02%, substantially outperforming smaller models and confirming the importance of scale in handling complex clinical text. Crucially, we identified a highly complementary error profile: the LLM excelled at syntactic consistency (e.g., date/numeric formatting) where humans faltered, while human experts provided superior semantic and contextual interpretation. We also found that the LLM's semantic errors were fundamental limitations that could not be mitigated by simple prompt engineering. These findings strongly support a human-in-the-loop (HITL) workflow in which the on-premise LLM serves as a collaborative tool, not a full replacement. It automates routine structuring and flags potential human errors, enabling imaging specialists to focus on high-level semantic validation. We discuss implications for structured reporting and interactive AI systems in clinical practice.

cs.HC↗

arXiv2026

CLOC: Contrastive Learning for Ordinal Classification with Multi-Margin N-pair Loss

In ordinal classification, misclassifying neighboring ranks is common, yet the consequences of these errors are not the same. For example, misclassifying benign tumor categories is less consequential, compared to an error at the pre-cancerous to cancerous threshold, which could profoundly influence treatment choices. Despite this, existing ordinal classification methods do not account for the varying importance of these margins, treating all neighboring classes as equally significant. To address this limitation, we propose CLOC, a new margin-based contrastive learning method for ordinal classification that learns an ordered representation based on the optimization of multiple margins with a novel multi-margin n-pair loss (MMNP). CLOC enables flexible decision boundaries across key adjacent categories, facilitating smooth transitions between classes and reducing the risk of overfitting to biases present in the training data. We provide empirical discussion regarding the properties of MMNP and show experimental results on five real-world image datasets (Adience, Historical Colour Image Dating, Knee Osteoarthritis, Indian Diabetic Retinopathy Image, and Breast Carcinoma Subtyping) and one synthetic dataset simulating clinical decision bias. Our results demonstrate that CLOC outperforms existing ordinal classification methods and show the interpretability and controllability of CLOC in learning meaningful, ordered representations that align with clinical and practical needs.

cs.CV↗

arXiv2026

L2CU: Learning to Complement Unseen Users

Recent research highlights the potential of machine learning models to learn to complement (L2C) human strengths; however, generalizing this capability to unseen users remains a significant challenge. Existing L2C methods oversimplify interaction between human and AI by relying on a single, global user model that neglects individual user variability, leading to suboptimal cooperative performance. Addressing this, we introduce L2CU, a novel L2C framework for human-AI cooperative classification with unseen users. Given sparse and noisy user annotations, L2CU identifies representative annotator profiles capturing distinct labeling patterns. By matching unseen users to these profiles, L2CU leverages profile-specific models to complement the user and achieve superior joint accuracy. We evaluate L2CU on datasets (CIFAR-10N, CIFAR-10H, Fashion-MNIST-H, Chaoyang and AgNews), demonstrating its effectiveness as a model-agnostic solution for improving human-AI cooperative classification.

cs.LG↗

arXiv2025

OUGS: Active View Selection via Object-aware Uncertainty Estimation in 3DGS

Recent advances in 3D Gaussian Splatting (3DGS) have achieved state-of-the-art results for novel view synthesis. However, efficiently capturing high-fidelity reconstructions of specific objects within complex scenes remains a significant challenge. A key limitation of existing active reconstruction methods is their reliance on scene-level uncertainty metrics, which are often biased by irrelevant background clutter and lead to inefficient view selection for object-centric tasks. We present OUGS, a novel framework that addresses this challenge with a more principled, physically-grounded uncertainty formulation for 3DGS. Our core innovation is to derive uncertainty directly from the explicit physical parameters of the 3D Gaussian primitives (e.g., position, scale, rotation). By propagating the covariance of these parameters through the rendering Jacobian, we establish a highly interpretable uncertainty model. This foundation allows us to then seamlessly integrate semantic segmentation masks to produce a targeted, object-aware uncertainty score that effectively disentangles the object from its environment. This allows for a more effective active view selection strategy that prioritizes views critical to improving object fidelity. Experimental evaluations on public datasets demonstrate that our approach significantly improves the efficiency of the 3DGS reconstruction process and achieves higher quality for targeted objects compared to existing state-of-the-art methods, while also serving as a robust uncertainty estimator for the global scene.

cs.CV↗

arXiv2025

Learning To Defer To A Population With Limited Demonstrations

This paper addresses the critical data scarcity that hinders the practical deployment of learning to defer (L2D) systems to the population. We introduce a context-aware, semi-supervised framework that uses meta-learning to generate expert-specific embeddings from only a few demonstrations. We demonstrate the efficacy of a dual-purpose mechanism, where these embeddings are used first to generate a large corpus of pseudo-labels for training, and subsequently to enable on-the-fly adaptation to new experts at test-time. The experiment results on three different datasets confirm that a model trained on these synthetic labels rapidly approaches oracle-level performance, validating the data efficiency of our approach. By resolving a key training bottleneck, this work makes adaptive L2D systems more practical and scalable, paving the way for human-AI collaboration in real-world environments. To facilitate reproducibility and address implementation details not covered in the main text, we provide our source code and training configurations at https://github.com/nil123532/learning-to-defer-to-a-population-with-limited-demonstrations.

cs.HC↗

arXiv2025

Context-AI Tunes: Context-Aware AI-Generated Music for Stress Reduction

Music plays a critical role in emotional regulation and stress relief; however, individuals often need different types of music tailored to their unique stress levels or surrounding environment. Choosing the right music can be challenging due to the overwhelming number of options and the time-consuming trial-and-error process. To address this, we propose Context-AI Tune (CAT), a system that generates personalized music based on environmental inputs and the user's self-assessed stress level. A 2x2 within-subject experiment (N=26) was conducted with two independent variables: AI (AI, NoAI) and Environment (Busy Hub, Quiet Library). CAT's effectiveness in reducing stress was evaluated using the Visual Analog Scale for Stress (VAS-S). Results show that CAT is more effective than manually chosen music in reducing stress by adapting to user context.

cs.HC↗

arXiv2025

SnapNCode: An Integrated Development Environment for Programming Physical Objects Interactions

Spatial computing technologies have the potential to revolutionize how we interact with the world around us. However, most modern integrated development environments (IDEs) have not fully adapted to this paradigm shift. For example, physical 3D objects in the real world are still represented as 2D text variables in code, creating a significant perceptual distance between these representations. In response to this challenge, we introduce SnapNCode, a novel IDE for spatial programming. SnapNCode enables programmers to capture various states of physical objects through live video streams from cameras and directly insert these visual representations into their code. Moreover, users can augment physical objects by attaching code snippets onto objects, which are opportunistically triggered when observed by cameras. We conducted a user study (N=12) to assess the usability of SnapNCode. Feedback from participants indicates that the system is easy-to-use and holds promise for daily casual uses and integration into a broader range of workflows.

cs.HC↗

arXiv2025

Toward a Human-Centered AI-assisted Colonoscopy System in Australia

While AI-assisted colonoscopy promises improved colorectal cancer screening, its success relies on effective integration into clinical practice, not just algorithmic accuracy. This paper, based on an Australian field study (observations and gastroenterologist interviews), highlights a critical disconnect: current development prioritizes machine learning model performance, overlooking essential aspects of user interface design, workflow integration, and overall user experience. Industry interactions reveal a similar emphasis on data and algorithms. To realize AI's full potential, the HCI community must champion user-centered design, ensuring these systems are usable, support endoscopist expertise, and enhance patient outcomes.

cs.HC↗

arXiv2024

Segment Beyond View: Handling Partially Missing Modality for Audio-Visual Semantic Segmentation

Augmented Reality (AR) devices, emerging as prominent mobile interaction platforms, face challenges in user safety, particularly concerning oncoming vehicles. While some solutions leverage onboard camera arrays, these cameras often have limited field-of-view (FoV) with front or downward perspectives. Addressing this, we propose a new out-of-view semantic segmentation task and Segment Beyond View (SBV), a novel audio-visual semantic segmentation method. SBV supplements the visual modality, which miss the information beyond FoV, with the auditory information using a teacher-student distillation model (Omni2Ego). The model consists of a vision teacher utilising panoramic information, an auditory teacher with 8-channel audio, and an audio-visual student that takes views with limited FoV and binaural audio as input and produce semantic segmentation for objects outside FoV. SBV outperforms existing models in comparative evaluations and shows a consistent performance across varying FoV ranges and in monaural audio settings.

cs.CV↗

arXiv2023

WebVLN: Vision-and-Language Navigation on Websites

Vision-and-Language Navigation (VLN) task aims to enable AI agents to accurately understand and follow natural language instructions to navigate through real-world environments, ultimately reaching specific target locations. We recognise a promising opportunity to extend VLN to a comparable navigation task that holds substantial significance in our daily lives, albeit within the virtual realm: navigating websites on the Internet. This paper proposes a new task named Vision-and-Language Navigation on Websites (WebVLN), where we use question-based instructions to train an agent, emulating how users naturally browse websites. Unlike the existing VLN task that only pays attention to vision and instruction (language), the WebVLN agent further considers underlying web-specific content like HTML, which could not be seen on the rendered web pages yet contains rich visual and textual information. Toward this goal, we contribute a dataset, WebVLN-v1, and introduce a novel approach called Website-aware VLN Network (WebVLN-Net), which is built upon the foundation of state-of-the-art VLN techniques. Experimental results show that WebVLN-Net outperforms current VLN and web-related navigation methods. We believe that the introduction of the new WebVLN task and its dataset will establish a new dimension within the VLN domain and contribute to the broader vision-and-language research community. The code is available at: https://github.com/WebVLN/WebVLN.

cs.CV↗

arXiv2023

An EEG-based Experiment on VR Sickness and Postural Instability While Walking in Virtual Environments

Previous studies showed that natural walking reduces the susceptibility to VR sickness. However, many users still experience VR sickness when wearing VR headsets that allow free walking in room-scale spaces. This paper studies VR sickness and postural instability while the user walks in an immersive virtual environment using an electroencephalogram (EEG) headset and a full-body motion capture system. The experiment induced VR sickness by gradually increasing the translation gain beyond the user's detection threshold. A between-group comparison between participants with and without VR sickness symptoms found some significant differences in postural stability but found none on gait patterns during the walking. In the EEG analysis, the group with VR sickness showed a reduction of alpha power, a phenomenon previously linked to a higher workload and efforts to maintain postural control. In contrast, the group without VR sickness exhibited brain activities linked to fine cognitive-motor control. The EEG result provides new insights into the postural instability theory: participants with VR sickness could maintain their postural stability at the cost of a higher cognitive workload. Our result also indicates that the analysis of lower-frequency power could complement behavioural data for continuous VR sickness detection in both stationary and mobile VR setups.

cs.HC↗

arXiv2022

Toward a Human-Centered AI-assisted Colonoscopy System

AI-assisted colonoscopy has received lots of attention in the last decade. Several randomised clinical trials in the previous two years showed exciting results of the improving detection rate of polyps. However, current commercial AI-assisted colonoscopy systems focus on providing visual assistance for detecting polyps during colonoscopy. There is a lack of understanding of the needs of gastroenterologists and the usability issues of these systems. This paper aims to introduce the recent development and deployment of commercial AI-assisted colonoscopy systems to the HCI community, identify gaps between the expectation of the clinicians and the capabilities of the commercial systems, and highlight some unique challenges in Australia.

cs.HC↗

arXiv2021

Identification of EEG Dynamics During Freezing of Gait and Voluntary Stopping in Patients with Parkinson's Disease

Mobility is severely impacted in patients with Parkinson's disease (PD), especially when they experience involuntary stopping from the freezing of gait (FOG). Understanding the neurophysiological difference between "voluntary stopping" and "involuntary stopping" caused by FOG is vital for the detection and potential intervention of FOG in the daily lives of patients. This study characterised the electroencephalographic (EEG) signature associated with FOG in contrast to voluntary stopping. The protocol consisted of a timed up-and-go (TUG) task and an additional TUG task with a voluntary stopping component, where participants reacted to verbal "stop" and "walk" instructions by voluntarily stopping or walking. Event-related spectral perturbation (ERSP) analysis was used to study the dynamics of the EEG spectra induced by different walking phases, which included normal walking, voluntary stopping and episodes of involuntary stopping (FOG), as well as the transition windows between normal walking and voluntary stopping or FOG. These results demonstrate for the first time that the EEG signal during the transition from walking to voluntary stopping is distinguishable from that of the transition to involuntary stopping caused by FOG. The EEG signature of voluntary stopping exhibits a significantly decreased power spectrum compared to that of FOG episodes, with distinctly different patterns in the delta and low-beta power in the central area. These findings suggest the possibility of a practical EEG-based treatment strategy that can accurately predict FOG episodes, excluding the potential confound of voluntary stopping.

physics.med-ph↗