Youngjo Min

I am a Ph.D. candidate in Computer Science at Duke University, co-advised by Professors David Carlson and Carlo Tomasi. My research focuses on 3D and 4D computer vision, multimodal learning, and autonomous driving.

I develop models for reconstructing and understanding dynamic environments from images, videos, and multimodal sensor data. My work spans 3D reconstruction, scene understanding, and sensor fusion, with applications ranging from human and animal modeling to autonomous driving.

More broadly, I aim to advance spatial intelligence by learning representations that capture the geometry, appearance, and motion of the physical world.

AI/ML Engineer Intern

05/2026 – 08/2026          Cruise (General Motors Autonomous Driving)

• Developed and applied feed-forward 3D Gaussian splatting models to reconstruct dynamic scenes from internal real-world driving logs, contributing to world model development for autonomous driving perception.

Graduate Research Assistant 

09/2023 - 02/2028 (expected)          Duke University           

• Develop 3D and 4D vision models for reconstructing and understanding dynamic scenes, with a focus on geometry, appearance, and motion.

• Develop multimodal perception models for autonomous driving, integrating information across sensors to support robust scene understanding in diverse environments.

• Develop methods for 3D human and animal modeling, including shape reconstruction, pose estimation, and motion analysis.

Researcher

09/2022 - 08/2023          Seoul National University

• Developed unsupervised open-set domain adaptation model using a vision-language model

Undergraduate Researcher

08/2020 - 08/2022          Korea University

• Developed a semi-supervised classification framework using a learnable confidence estimator and a confidence-aware consistency regularization.

• Created an implicit neural representation-based style transfer model that can handle a variety of flexible image translation tasks.

• Developed a framework for 3D human recovery from a single image by combining meta-learning with test-time optimization.

• Devised a semantic matching model that jointly optimizes feature extraction and cost aggregation modules

Air Force Sergeant

02/2017 - 02/2019          Republic of Korea Air Force 

• Participated in multiple Korea–U.S. joint military exercises and operations conducted with U.S. Forces Korea and the U.S. 7th Air Force.

LiGRe: Grounding Feed-Forward Gaussian Reconstruction in LiDAR Geometry for Joint RGB and LiDAR Rendering

Youngjo Min

Under review

[Paper]

EviOcc: Preserving Visual Evidence for Cross-Dataset 3D Occupancy Anomaly Detection

Youngjo Min and David Carlson

Under review

[Paper]

Pose Splatter: A 3D Gaussian Splatting Model for Quantifying Animal Pose and Appearance

Jack Goffinet*, Youngjo Min*, Carlo Tomasi, and David Carlson 

(* Equal Contribution) 

Neural Information Processing Systems (NeurIPS), 2025.

[Paper]  [Project]

ConMatch: Semi-Supervised Learning with Confidence-Guided Consistency Regularization

Jiwon Kim*, Youngjo Min*, Daehwan Kim*, Gyuseong Lee, Junyoung Seo, Kwangrok Ryoo, and Seungryong Kim 

(* Equal Contribution)

European Conference on Computer Vision (ECCV), 2022.

[Paper]  [Project]

Controllable Style Transfer via Test-time Training of Implicit Neural Representation 

Sunwoo Kim*, Youngjo Min*, Younghun Jung*, and Seungryong Kim 

(* Equal Contribution) 

Pattern Recognition, 2024.

[Paper]  [Project]

UOTA: Unsupervised Open-Set Task Adaptation Using a Vision-Language Foundation Model

Youngjo Min*, Kwangrok Ryu*, Bumsoo Kim, and Taesup Kim 

(* Equal Contribution)

ICML 2023 Workshop on Efficient Systems for Foundation Models.

[Paper]

Joint Learning of Feature Extraction and Cost Aggregation for Semantic Correspondence

Jiwon Kim, Youngjo Min, Mira Kim, and Seungryong Kim 

IEEE International Conference on Acoustics, Speech, and Signal Processing

(ICASSP), 2022.

[Paper] 

Meta-Learned Initialization for 3D Human Recovery

Mira Kim, Youngjo Min, Jiwon Kim, and Seungryong Kim

IEEE International Conference on Image Processing (ICIP), 2022.

[Paper] 

Semi-Supervised Learning with Confidence

Youngjo Min, Kwangrok Ryoo, Jiwon Kim, Junyoung Seo, Kyusung Lee, Seungryong Kim

IEIE Conference, 2022 (in Korean).

[Paper] 

Confidence-guided Semantic Correspondence

Youngjo Min, Kwangrok Ryoo, Jiwon Kim, Suhyung Choi, Kyusun Cho, Minhyek Jeon, Seungryong Kim

IEIE Conference, 2022 (in Korean).

[Paper]