Youn-Hee Gil
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Youn-Hee Gil has a total of 44 co-authors. Here are the top co-authors by number of shared publications:
- Heesook Shin
Electronics and Telecommunications Research Institute, Daejeon, Republic of Korea
8 shared publications - Seongmin Baek
Electronics and Telecommunications Research Institute, Daejeon, Republic of Korea
7 shared publications - Hyunjun Lim
Electronics and Telecommunications Research Institute, Daejeon, Republic of Korea
4 shared publications - Sungjin Hong
Electronics and Telecommunications Research Institute, Daejeon, Republic of Korea
4 shared publications - Yongho Lee
Content Research Laboratory, Electronics and Telecommunications Research Institute, Daejeon, Republic of Korea
4 shared publications - Cho-Rong Yu
Electronics and Telecommunications Research Institute, Daejeon, Republic of Korea
3 shared publications
Latest Publications
- Open Access
- 16 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-11
Decode the Avatar Face: Affect Recovery in Dyadic VR
- 16 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-11
Open AccessDecode the Avatar Face: Affect Recovery in Dyadic VR
- Open Access
- 16 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-3
Human Pose Estimation as an Online Metric Ruler for Egocentric SLAM
- 16 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-3
Open AccessHuman Pose Estimation as an Online Metric Ruler for Egocentric SLAM
- Open Access
- 16 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-2
EgoLiveHAND: Real-Time Egocentric 3D Hand Pose Estimation with Stereo Fisheye Glasses
- 16 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-2
Open AccessEgoLiveHAND: Real-Time Egocentric 3D Hand Pose Estimation with Stereo Fisheye Glasses
- Open Access
- 16 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-3
Real-Time Full-Body Avatar Control in VR Using Glasses-Mounted Stereo Cameras
- 16 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-3
Open AccessReal-Time Full-Body Avatar Control in VR Using Glasses-Mounted Stereo Cameras
- Open Access
- 16 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-2
Odometry-Free Cross-View Pose Fusion for XR Interaction via Visibility-Guided Alignment
- 16 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-2
Open AccessOdometry-Free Cross-View Pose Fusion for XR Interaction via Visibility-Guided Alignment
- Open Access
- 23 February 2026
- Published by Springer Nature in Virtual Reality
- Vol. 30 (2)
The effect of facial cue placement in virtual collaboration
Sharing facial expressions plays an important role in communicating emotions during interaction. However, in collaborative tasks with heavy visual and physical demands, it can be difficult to share or notice these expressions. Virtual Reality (VR) offers opportunities to overcome these limitations by enabling non-verbal cues to be positioned flexibly in the shared environment so they remain visible and salient. Yet, how such cues help in visually and physically demanding collaboration remains unclear. This research investigates how facial cue placement affects collaboration during virtual object interaction. We developed a VR prototype that shares facial expressions using a 3-Dimensional avatar head, placed on a collaborator’s hand or task-related objects. A user study based on 30 participants evaluated the impact of these placements showed that only cues placed “on hand" had increased participants’ gaze activities to where the placement was, whether consciously or automatically. Interestingly the absence of facial cues boost participants predictions toward their acting partner’s satisfaction level, suggesting hand cues may hinder emotional attribution. No clear performance differences were observed compared to a traditional VR setup. We discuss the design and practical implications of facial cue placement and future research directions for emotional awareness in VR collaboration.
- 23 February 2026
- Published by Springer Nature in Virtual Reality
- Vol. 30 (2)
Open AccessThe effect of facial cue placement in virtual collaboration
Sharing facial expressions plays an important role in communicating emotions during interaction. However, in collaborative tasks with heavy visual and physical demands, it can be difficult to share or notice these expressions. Virtual Reality (VR) offers opportunities to overcome these limitations by enabling non-verbal cues to be positioned flexibly in the shared environment so they remain visible and salient. Yet, how such cues help in visually and physically demanding collaboration remains unclear. This research investigates how facial cue placement affects collaboration during virtual object interaction. We developed a VR prototype that shares facial expressions using a 3-Dimensional avatar head, placed on a collaborator’s hand or task-related objects. A user study based on 30 participants evaluated the impact of these placements showed that only cues placed “on hand" had increased participants’ gaze activities to where the placement was, whether consciously or automatically. Interestingly the absence of facial cues boost participants predictions toward their acting partner’s satisfaction level, suggesting hand cues may hinder emotional attribution. No clear performance differences were observed compared to a traditional VR setup. We discuss the design and practical implications of facial cue placement and future research directions for emotional awareness in VR collaboration.
- Open Access
- 12 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 14, 8128-8148
Egocentric Hand Activity Video Dataset and Bidirectional Motion-Priors for Hand Action Recognition
Recognizing tool-based hand activities from a first-person view is a critical yet challenging task in computer vision, due to the complexity of hand-object interactions and often subtle, ambiguous motion patterns. In real-world manufacturing scenarios, these challenges are exacerbated by bidirectional action pairs whose visual cues are almost identical, with differences revealed only through subtle motion dynamics. However, existing datasets rarely capture these direction-sensitive interactions at scale, particularly in realistic tool-use contexts, limiting the ability of current models to learn fine-grained motion dynamics essential for accurate recognition. We introduce Ego-Bi (Egocentric-Bidirectional dataset), a large-scale, real-world egocentric RGB video dataset comprising 1,223 video sequences and 622,737 frames that cover diverse tool-use activities in unconstrained environments. Ego-Bi provides an extended 38-category hand type taxonomy, detailed object–tool labels, and challenging bidirectional action pairs, offering rich semantic and temporal cues for modeling complex hand–object interactions. In addition, to address the ambiguity in motion dynamics, we propose a BMP (Bidirectional Motion Prior module) that derives rotation and directional cues from predicted 3D hand poses to improve class separability of visually similar actions. Experimental results on Ego-Bi demonstrate that our approach improves bidirectional action recognition accuracy by +8.96% over the baseline, while also yielding consistent gains across general action classes without requiring costly 3D pose annotations. Furthermore, the proposed motion priors generalize effectively to other egocentric benchmarks, underscoring their robustness in handling visually similar, direction-sensitive actions.
- 12 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 14, 8128-8148
Open AccessEgocentric Hand Activity Video Dataset and Bidirectional Motion-Priors for Hand Action Recognition
Recognizing tool-based hand activities from a first-person view is a critical yet challenging task in computer vision, due to the complexity of hand-object interactions and often subtle, ambiguous motion patterns. In real-world manufacturing scenarios, these challenges are exacerbated by bidirectional action pairs whose visual cues are almost identical, with differences revealed only through subtle motion dynamics. However, existing datasets rarely capture these direction-sensitive interactions at scale, particularly in realistic tool-use contexts, limiting the ability of current models to learn fine-grained motion dynamics essential for accurate recognition. We introduce Ego-Bi (Egocentric-Bidirectional dataset), a large-scale, real-world egocentric RGB video dataset comprising 1,223 video sequences and 622,737 frames that cover diverse tool-use activities in unconstrained environments. Ego-Bi provides an extended 38-category hand type taxonomy, detailed object–tool labels, and challenging bidirectional action pairs, offering rich semantic and temporal cues for modeling complex hand–object interactions. In addition, to address the ambiguity in motion dynamics, we propose a BMP (Bidirectional Motion Prior module) that derives rotation and directional cues from predicted 3D hand poses to improve class separability of visually similar actions. Experimental results on Ego-Bi demonstrate that our approach improves bidirectional action recognition accuracy by +8.96% over the baseline, while also yielding consistent gains across general action classes without requiring costly 3D pose annotations. Furthermore, the proposed motion priors generalize effectively to other egocentric benchmarks, underscoring their robustness in handling visually similar, direction-sensitive actions.
- Open Access
- 14 December 2025
- Published by Association for Computing Machinery (ACM)
- p. 1-3
Egocentric 3D Pose Estimation with Temporally Guided Pose Refinement
- 14 December 2025
- Published by Association for Computing Machinery (ACM)
- p. 1-3
Open AccessEgocentric 3D Pose Estimation with Temporally Guided Pose Refinement
- Open Access
- 12 November 2025
- Published by Association for Computing Machinery (ACM)
- p. 1-2
StereoVisPoseNet: Stereo-based Visibility-aware Egocentric 3D Pose Estimation Network
- 12 November 2025
- Published by Association for Computing Machinery (ACM)
- p. 1-2
Open AccessStereoVisPoseNet: Stereo-based Visibility-aware Egocentric 3D Pose Estimation Network
- Open Access
- 12 October 2025
- Published by Association for Computing Machinery (ACM)
- p. 15-19
A Novel HMD-Mounted Contactless Proxi-rPPG Sensor for Emotion Recognition in Free-Moving VR Environments
- 12 October 2025
- Published by Association for Computing Machinery (ACM)
- p. 15-19
Open AccessA Novel HMD-Mounted Contactless Proxi-rPPG Sensor for Emotion Recognition in Free-Moving VR Environments