APSIPA ASC 2026
ATW-GS
Anisotropic transmittance windows for anti-aliased 3D Gaussian splatting
- Anti-aliasing
- Multi-scale rendering
- 3D Gaussian splatting
PhD Candidate · The Hong Kong Polytechnic University
Hana L. Goshu
Reliable AI for 3D visual perception and scene understanding
I develop methods that recover and understand three-dimensional scenes from ordinary images and video, even when the views are few, uneven, or limited by how the camera moved. My aim is AI that is accurate and efficient, and honest about how much it really knows.
- Peer-reviewed publications
- 10
- Courses taught
- 6
Research focus
Seeing the world in three dimensions, reliably
One question runs through my work: what can be recovered about a three-dimensional scene when the images available do not determine the answer? Dense, well-distributed capture is a laboratory condition. The settings that matter most, such as a camera on a vehicle, a drone following a flight path, or a handheld capture of a space walked through once, fix the coverage in advance.
Reconstruction from sparse views
Recovering geometry and appearance from a few, unevenly distributed views, using stereo priors, targeted densification and optimization that stays stable when the data are insufficient.
3D detection and scene understanding
Recognizing and locating objects inside reconstructed scenes, with volumetric learning that stays robust when views are sparse and noisy.
Efficient, anti-aliased rendering
Reducing model size and training cost without losing quality, and keeping rendering stable at scales and distances outside the fitted regime.
Trustworthy AI and uncertainty
Physics-informed learning with calibrated uncertainty for engineering prediction, and models that report where their own evidence is thin.
Selected systems
Research systems
SMAGaussian
Stabilizing joint optimization with weight averaging and multi-resolution supervision for egocentric 3D Gaussian splatting
- 3D Gaussian splatting
- Egocentric video
- Optimization
FastGS-Adaptive
Structure-aware, view-consistent densification for accelerated 3D Gaussian splatting
- Efficiency
- Densification
- 3D Gaussian splatting
Publications
Selected publications
Journal article2026
NeRF-UAVeL: Unified Attention-Driven Volumetric Learning for Robust NeRF-Based 3D Object Detection
H. L. Goshu, T. Wakjira, M. Atlaw, K.-C. Chan, S. Lai, K.-M. Lam
Neurocomputing
Journal article2026
Physics-Informed Transformer with Token Classification for Hysteretic Response Prediction and Uncertainty Quantification of Reinforced Concrete Columns
T. Wakjira, H. L. Goshu
Applied Soft Computing
Conference paper2024CORE A*
Point Cloud Densification for 3D Gaussian Splatting from Sparse Input Views
K.-C. Chan, J. Xiao, H. L. Goshu, K.-M. Lam
ACM MM 2024
News
Latest updates
Talk
Invited seminar at Boğaziçi University
I gave an invited seminar at the Institute for Data Science and Artificial Intelligence (DSAI), Boğaziçi University, titled “Toward Reliable AI for 3D Visual Perception and Scene Understanding.”
Read moreTalks
All talks- 17 Sep 2026
Toward Reliable AI for 3D Visual Perception and Scene Understanding
Invited seminar, Institute for Data Science and Artificial Intelligence, Boğaziçi University
- 17 Jul 2025
Feature Calibration Mechanisms for NeRF-Based 3D Object Detection
PolyU EEE Research Student Conference (PRSC 2025)
- Dec 2024
NeRF-FCM: Feature Calibration Mechanisms for NeRF-Based 3D Object Detection
APSIPA Annual Summit and Conference 2024
Teaching
Teaching6 courses at undergraduate and postgraduate level, taught in English.
- Pattern Recognition: Theory and Applications
- Digital Image Processing
- Computer Programming
- Web Systems and Technologies
- Artificial Intelligence and Science Fiction
- Computational Intelligence and Its Applications
Get in touch
Collaborations, talks and teaching
I welcome conversations about research collaboration, invited talks and teaching in AI and computer vision.