About

Jin-Hyuk Kim is an Associate Professor in the Department of Economics at the University of Colorado Boulder. His research interests include the economic analysis of intellectual property rights, especially copyrights, relational contracts, team organizations, post-employment restrictions, lobbying, and political institutions. His current research focuses on the impact of international harmonization of copyright levies on social welfare. He holds a PhD and an MA from Cornell University, and a BA from Yonsei University. His academic background and research focus on various fields within economics, including Industrial Organization, Organization Economics, and Political Economics. Jin-Hyuk Kim is actively engaged in exploring issues related to the digital economy, relational contracts, entrepreneurship, and antitrust, contributing to the understanding of how legal and institutional frameworks influence economic outcomes.

Research topics

• Computer Science
• Artificial Intelligence
• Engineering
• Human–computer interaction
• Mechanical engineering
• Engineering drawing
• Materials science
• Computer hardware
• Aesthetics
• Software engineering
• Operating system
• Manufacturing engineering
• Multimedia
• World Wide Web
• Art
• Visual arts

Selected publications

• End-to-end architectural framework for proximity-based preemptive haptic feedback in safe excavation

  Automation in Construction · 2026-04-18

  article
  PublisherDOI

• LASOR: Towards Clinically Transparent and Explainable Ophthalmic Report Generation via Lesion-Aware Segmentation

  2026-03-06

  article

  Automated ophthalmic report generation aims to reduce the diagnostic burden on retinal specialists by producing clinically accurate and standardized descriptions from medical imaging. However, current research predominantly remains fundus-centric and rarely exploits OCT-derived spatial evidence, limiting clinical transparency by obscuring which anatomical regions drive diagnostic decisions. To address these limitations, we propose LASOR (Lesion-Aware Segmentation-Guided Ophthalmic Report Generation), which extracts multi-scale features to robustly capture both small focal abnormalities and broader anatomical structures, generating reliable segmentation masks as spatial priors for report generation. Specifically, we utilize a lesion-aware patch weighting module to emphasize abnormal regions and leverage a curated instruction dataset incorporating spatial mask information to enhance the diagnostic capabilities of the proposed model. In addition, we introduce a mask-guided cross-modal consistency loss that strengthens vision-language alignment between pathological regions and their diagnostic descriptions. Extensive experiments on a retinal OCT dataset that includes twenty pathological conditions exhibit state-of-the-art performance, underscoring LASOR’s potential to advance clinically transparent ophthalmic report generation systems.

  PublisherDOI

• EUREXA: End-User Reconfiguration of Environment with eXplainable Augmentation for Generative Fabrication

  2026-04-13 · 1 citations

  articleOpen accessSenior author

  Augmentation allows rapid reconfiguration of passive physical interfaces to improve accessibility, support independent living through domestic automation, and more. However, its potential is largely unrealized for novice users due to several key barriers. First, users rarely identify latent interaction problems within their built environments. Second, they often lack the knowledge to clearly express design intent. Third, many innovative solutions remain in research prototypes, limiting access.

  PublisherDOI

• LumosX: 3D Printed Anisotropic Light-Transfer

  2025-04-24 · 2 citations

  articleOpen accessSenior author
  PublisherDOI

• A high-throughput, fully automated competition assay to evaluate SARS-CoV-2 neutralizing responses and epitope specificity in clinical samples

  Scientific Reports · 2025-04-04 · 1 citations

  articleOpen access

  Coronavirus disease-2019 (COVID-19) remains a critical global health concern. We developed a fully automated, high-throughput competition immunoassay to elucidate how epitope recognition on the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) spike receptor-binding domain (RBD) correlates with neutralizing activity. Analysis of clinical samples from both SARS-CoV-2-infected and vaccinated individuals revealed that vaccination elicits significantly higher antibody titers across multiple S1 subunit epitopes compared to natural infection. Notably, median antibody levels against the receptor-binding motif (RBM) exceeded 50% in both cohorts, highlighting the RBM as a key target for antibody induction irrespective of immune origin. Furthermore, the strongest correlation with neutralizing activity was observed for antibodies directed against the broader S1 subunit, indicating that epitopes outside the RBM also contribute to neutralization. These findings underscore the importance of both RBM- and non-RBM-directed antibodies in effective immune defense against SARS-CoV-2. Our assay enables large-scale, reliable quantification of neutralizing antibodies and provides critical insights for developing improved diagnostic antigens and vaccine strategies aimed at eliciting robust, multi-epitope immune responses.

  PublisherOA PDFDOI

• Stretchable Micro-Wrinkled Carbon Nanotube-Assembled Skin-Adhesive Patches with Suction-Cup Patterns for Human Breath-Derived Moisture Energy Harvesting

  ACS Nano · 2025-05-26 · 7 citations

  article

  With significant advances in self-powered, stretchable, and skin-attachable electronics, harvesting energy from ubiquitous moisture has emerged as a promising method for powering wearable and adhesive devices. However, current moisture energy harvesting (MEH) devices still face challenges in direct application to skin surfaces, mainly due to insufficient stretchability and weak adhesion, particularly under wet conditions. Here, we construct a stretchable and skin-adhesive MEH patch by harnessing microwrinkled carbon nanotube (CNT) sheets featuring asymmetric oxygen content and a highly elastic silicone rubber-polymer substrate with suction-cup patterns (SP). The developed MEH patch (2 cm × 4 cm) achieves an open-circuit voltage of ∼102 mV and a short-circuit current of ∼1.75 mA/m2 under ambient humidity variations. Notably, it maintains stable electrical output even when stretched up to 300% strain. The SP architecture introduced in the patch ensures robust adhesion to both dry and wet skin surfaces with the application of preload. Consequently, the stretchable and adhesive MEH patch can effectively convert breath-induced moisture energy into electric output on the philtrum, enabling self-powered monitoring of various respiratory patterns.

  PublisherDOI

• Solving Instance Detection from an Open-World Perspective

  2025-06-10 · 2 citations

  article

  Instance detection (InsDet) aims to localize specific object instances within a novel scene imagery based on given visual references. Technically, it requires proposal detection to identify all possible object instances, followed by instance-level matching to pinpoint the ones of interest. Its open-world nature supports its broad applications from robotics to AR/VR but also presents significant challenges: methods must generalize to unknown testing data distributions because (1) the testing scene imagery is unseen during training, and (2) there are domain gaps between visual references and detected proposals. Existing methods tackle these challenges by synthesizing diverse training examples or utilizing off-the-shelf foundation models (FMs). However, they only partially capitalize the available open-world information. In contrast, we approach InsDet from an Open-World perspective, introducing our method IDOW. We find that, while pretrained FMs yield high recall in instance detection, they are not specifically optimized for instance-level feature matching. Therefore, we adapt pretrained FMs for improved instance-level matching using open-world data. Our approach incorporates metric learning along with novel data augmentations, which sample distractors as negative examples and synthesize novel-view instances to enrich the visual references. Extensive experiments demonstrate that our method significantly outperforms prior works, achieving >10 AP over previous results on two recently released challenging benchmark datasets in both conventional and novel instance detection settings.

  PublisherDOI

• FABRIC: FAbricating Bodily-Expressive Robots for Inclusive and Low-Cost Design

  2025-10-19

  articleSenior author

  Sign language serves individuals with hearing impairments as a crucial communication mode operating through visual-manual means. While there has been established theory and agreement about embodiment in multiple fields, only limited research has deeply engaged to lower access to the physical body for spatial perception and engagement. Embodied robots are often cost-prohibitive, and existing open-source robot fabrication packages are limited in their ability to fully address communication nuances, typically running only on predefined programs. Reprogramming for broader bodily interactions, such as gestures in various domains (e.g., construction), is nearly impossible unless expertise precedes. We introduce FABRIC, an end-to-end toolkit for fabricating and programming bodily language for unique human-robot interactions. The toolkit includes a fully 3D-printable robot, designed for consumer-grade FDM machinery, that learns from demonstration (LfD) to capture and translate users’ bodily expressions through its upper torso (arms and hands) movements. A visual programming interface enables appending or sequencing demonstrations from various sources, i.e., videos, cameras, and expandable word/phrase/sentence libraries.

  PublisherDOI

• Cholesterol Uptake Capacity as a Prognostic Marker of Cardiovascular Events for Patients with Coronary Artery Disease

  Journal of Atherosclerosis and Thrombosis · 2025-03-28

  articleOpen access

  AIM: Cholesterol uptake capacity (CUC) is a functional assessment of high-density lipoprotein (HDL) and has drawn attention for the risk stratification of atherosclerotic cardiovascular disease (ASCVD). This study evaluated the usefulness of HDL-CUC as a predictive marker for long-term ASCVD events in patients with coronary artery disease (CAD). METHODS: This retrospective observational study included 503 patients with CAD who underwent coronary revascularization. Blood was sampled from the participants within three months before or after index revascularization. The CUC was assayed using a previously reported automated system. The study population was divided into three groups according to the tertiles of CUC levels. The primary outcome was ASCVD events, which were defined as a composite of all-cause death, acute myocardial infarction, stroke, and peripheral artery disease. RESULTS: A total of 29 events were observed during the follow-up (median 2.8 years). The risk of the primary outcome in the low-CUC group was significantly higher than that in the high-CUC group (3-year incidence: low CUC 8.8% vs. high CUC 4.0%; log-rank p = 0.046). After adjusting for age and sex, the risk in the low-CUC group relative to that in the high-CUC group remained significantly high (hazard ratio 3.17, 95% confidence interval 1.05-9.54, p = 0.040). CONCLUSION: Low CUC in patients with CAD were associated with a higher risk of ASCVD events after coronary revascularization than high CUC levels. The assessment of HDL functionality measured by CUC would be useful for the risk prediction of ASCVD after coronary revascularization.

  PublisherOA PDFDOI

• Solving Instance Detection from an Open-World Perspective

  ArXiv.org · 2025-03-01

  preprintOpen access

  Instance detection (InsDet) aims to localize specific object instances within a novel scene imagery based on given visual references. Technically, it requires proposal detection to identify all possible object instances, followed by instance-level matching to pinpoint the ones of interest. Its open-world nature supports its broad applications from robotics to AR/VR but also presents significant challenges: methods must generalize to unknown testing data distributions because (1) the testing scene imagery is unseen during training, and (2) there are domain gaps between visual references and detected proposals. Existing methods tackle these challenges by synthesizing diverse training examples or utilizing off-the-shelf foundation models (FMs). However, they only partially capitalize the available open-world information. In contrast, we approach InsDet from an Open-World perspective, introducing our method IDOW. We find that, while pretrained FMs yield high recall in instance detection, they are not specifically optimized for instance-level feature matching. Therefore, we adapt pretrained FMs for improved instance-level matching using open-world data. Our approach incorporates metric learning along with novel data augmentations, which sample distractors as negative examples and synthesize novel-view instances to enrich the visual references. Extensive experiments demonstrate that our method significantly outperforms prior works, achieving >10 AP over previous results on two recently released challenging benchmark datasets in both conventional and novel instance detection settings.

  PublisherOA PDFDOI

Recent grants

• Collaborative Research: HCC: Small: Programmable Visual Capabilities of Environments through 3D printed Light-transfer

  NSF · $400k · 2022–2026

• FW-HTF-P: Upskilling Craftspeople to Prepare for the Future of End-user Driven Manufacturing

  NSF · $150k · 2022–2024

Frequent coauthors

• Tom Yeh

  University of Colorado Boulder

  17 shared

• Haruki Takahashi

  Tohoku Institute of Technology

  9 shared

• Himani Deshpande

  Texas A&M University

  9 shared

• Sungjoo Lee

  8 shared

• Clement Zheng

  National University of Singapore

  8 shared

• Xiang Chen

  University of California, Los Angeles

  7 shared

• Jennifer Mankoff

  University of Washington

  6 shared

• Nahyun Kwon

  Texas A&M University

  6 shared

Education

• Ph.D.

  Cornell University

• M.A.

  Cornell University

• B.A.

  Yonsei University