Alex Hanson

I am a Postdoctoral Fellow at the University of Toronto, working with Prof. David Lindell.

I received my Ph.D. in Computer Science from the University of Maryland, where I was advised by Prof. Tom Goldstein. I also hold an M.Sc. in Computer Science from the University of Maryland and dual B.Sc. degrees in Mathematics and Mechanical Engineering from the University of Illinois.

My research interests span computer vision, computer graphics, and machine learning, with a focus on 3D reconstruction and scene representations.

Previously, I interned at Amazon Studios and was awarded the NDSEG Graduate Fellowship.

           

Research
RIDE-Splat: Reflection Removal for In-the-Wild Driving Environments
Kevin David Hayes, Alex Hanson, Allen Tu, Pranav Asthana, Matthias Zwicker, Tom Goldstein
arXiv 2026
Reconstruct reflection-free 3D Gaussian Splatting scenes from dashcam footage by separating the world-consistent road scene from camera-attached windshield reflections.
Project Page | arXiv | Code | Dataset
SplatSuRe: Selective Super-Resolution for Multi-view Consistent 3D Gaussian Splatting
Pranav Asthana, Alex Hanson, Allen Tu, Tom Goldstein, Matthias Zwicker, Amitabh Varshney
CVPR 2026
We enhance 3D Gaussian Splatting by selectively injecting super-resolution only where high-frequency detail is missing, yielding sharper results and improved perceptual quality without introducing multi-view inconsistencies.
Project Page | arXiv | Code
Speedy Deformable 3D Gaussian Splatting: Fast Rendering and Compression of Dynamic Scenes
Allen Tu*, Haiyang Ying*, Alex Hanson, Yonghan Lee, Tom Goldstein, Matthias Zwicker
CVPR 2026
We significantly accelerate Deformable 3D Gaussian Splatting while preserving high video quality through Gaussian pruning and trajectory grouping.
Project Page | arXiv | Code
Speedy-Splat: Fast 3D Gaussian Splatting with Sparse Pixels and Sparse Primitives
Alex Hanson, Allen Tu, Geng Lin, Vasu Singla, Matthias Zwicker, Tom Goldstein
CVPR 2025
We address two inefficiencies in the 3D-GS rendering pipeline by introducing a more efficient Gaussian localization algorithm and a novel training-time pruning technique.
Project Page | arXiv | Code
PUP 3D-GS: Principled Uncertainty Pruning for 3D Gaussian Splatting
Alex Hanson*, Allen Tu*, Vasu Singla, Mayuka Jayawardhana, Matthias Zwicker, Tom Goldstein
CVPR 2025
We find that the sensitivity of a Gaussian to the L2 loss across the scene -- captured as as a second order approximation -- is a surprisingly effective pruning score.
Project Page | arXiv | Code
Rethinking Common Assumptions To Mitigate Racial Bias in Face Recognition Datasets
Matthew Gwilliam, Srinidhi Hegde, Lade Tinubu, Alex Hanson
ICCV Workshops 2021 - HTCV Workshop Best Paper Runner-Up
Training on some skewed race distributions of a popular training dataset to mitigate racial bias produces more equitable models.
arXiv | Code
SVMax: A Feature Embedding Regularizer
Ahmed Taha, Alex Hanson, Abhinav Shrivastava, Larry Davis
arXiv 2021
Spreading out feature embeddings via SVD improves image retreival and mitigates model collapse.
Project Page | arXiv | Code
Bidirectional Convolutional LSTM for the Detection of Violence in Videos
Alex Hanson*, Koutilya PNVR*, Sanjukta Krishnagopal, Larry Davis
ECCV Workshops 2018
Two new architectures for detecting violence in videos from differnt violence domains.
Paper | Code | Poster
Extra

I enjoy creating and solving puzzles in my free time, including escape room / puzzle hunt type puzzles.

Here are two of my puzzles that have appeared in past UMD puzzle hunts:
Harry Potter and The Giver.

Source Code

Style and layout credit