Chunghyun Park

I am a Ph.D. student at POSTECH where I am advised by Prof. Minsu Cho and a member of the Computer Vision Lab. Previously, I completed my M.S. in Artificial Intelligence and B.S. in Mechanical Engineering at POSTECH.

My research lies in machine learning and computer vision including, but not limited to, 3D scene understanding and reconstruction. I am particularly interested in 3D perception on point clouds for applications in AR/VR and robotics.

If you are interested in my research projects, please feel free to contact me.

News

Feb 27, 2025	Mosaic3D is accepted to CVPR 2025.
May 22, 2024	I am selected as an outstanding reviewer for CVPR 2024.
Feb 27, 2024	A paper about 3D semantic correspondence is accepted to CVPR 2024.
Dec 11, 2023	I joined NVIDIA Research Taiwan as a research intern.
Nov 2, 2023	I gave a spotlight talk at Korea AI Summit 2023.

Selected Publications

* indicates equal contribution.

Mosaic3D: Foundation Dataset and Model for Open-Vocabulary 3D Segmentation

Junha Lee*, Chunghyun Park*, Jaesung Choe, Yu-Chiang Frank Wang, Jan Kautz, Minsu Cho, and Chris Choy

IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2025

Abs arXiv Website

We tackle open-vocabulary 3D scene segmentation tasks by introducing a novel data generation pipeline and training framework. Our work targets three essential aspects required for an effective dataset: precise 3D region segmentation, comprehensive textual descriptions, and sufficient dataset scale. By leveraging state-of-the-art open-vocabulary image segmentation models and region-aware vision-language models (VLM), we develop an automatic pipeline capable of producing high-quality 3D mask-text pairs. Applying this pipeline to multiple 3D scene datasets, we create Mosaic3D-5.6M, a dataset of more than 30K annotated scenes with 5.6M mask-text pairs, significantly larger than existing datasets. Building on these data, we propose Mosaic3D, a 3D visiual foundation model (3D-VFM) combining a 3D encoder trained with contrastive learning and a lightweight mask decoder for open-vocabulary 3D semantic and instance segmentation. Our approach achieves state-of-the-art results on open-vocabulary 3D semantic and instance segmentation benchmarks including ScanNet200, Matterport3D, and ScanNet++, with ablation studies validating the effectiveness of our large-scale training data.
Learning SO(3)-Invariant Semantic Correspondence via Local Shape Transform

Chunghyun Park*, Seungwook Kim*, Jaesik Park, and Minsu Cho

IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2024
Presented in Workshop on Equivariant Vision: From Theory to Practice @ CVPR 2024
Abs arXiv Code Website

Establishing accurate 3D correspondences between shapes stands as a pivotal challenge with profound implications for computer vision and robotics. However, existing self-supervised methods for this problem assume perfect input shape alignment, restricting their real-world applicability. In this work, we introduce a novel self-supervised Rotation-Invariant 3D correspondence learner with Local Shape Transform, dubbed RIST, that learns to establish dense correspondences between shapes even under challenging intra-class variations and arbitrary orientations. Specifically, RIST learns to dynamically formulate an SO(3)-invariant local shape transform for each point, which maps the SO(3)-equivariant global shape descriptor of the input shape to a local shape descriptor. These local shape descriptors are provided as inputs to our decoder to facilitate point cloud self- and cross-reconstruction. Our proposed self-supervised training pipeline encourages semantically corresponding points from different shapes to be mapped to similar local shape descriptors, enabling RIST to establish dense point-wise correspondences. RIST demonstrates state-of-the-art performances on 3D part label transfer and semantic keypoint transfer given arbitrarily rotated point cloud pairs, outperforming existing methods by significant margins.
Stable and Consistent Prediction of 3D Characteristic Orientation via Invariant Residual Learning

Seungwook Kim*, Chunghyun Park*, Yoonwoo Jeong, Jaesik Park, and Minsu Cho

International Conference on Machine Learning (ICML), 2023

Abs arXiv

Learning to predict reliable characteristic orientations of 3D point clouds is an important yet challenging problem, as different point clouds of the same class may have largely varying appearances. In this work, we introduce a novel method to decouple the shape geometry and semantics of the input point cloud to achieve both stability and consistency. The proposed method integrates shape-geometry-based SO(3)-equivariant learning and shape-semantics-based SO(3)-invariant residual learning, where a final characteristic orientation is obtained by calibrating an SO(3)-equivariant orientation hypothesis using an SO(3)-invariant residual rotation. In experiments, the proposed method not only demonstrates superior stability and consistency but also exhibits state-of-the-art performances when applied to point cloud part segmentation, given randomly rotated inputs.
PointMixer: MLP-Mixer for Point Cloud Understanding

Jaesung Choe*, Chunghyun Park*, Francois Rameau, Jaesik Park, and In So Kweon

European Conference on Computer Vision (ECCV), 2022

Abs arXiv Code

MLP-Mixer has newly appeared as a new challenger against the realm of CNNs and Transformer. Despite its simplicity compared to Transformer, the concept of channel-mixing MLPs and token-mixing MLPs achieves noticeable performance in image recognition tasks. Unlike images, point clouds are inherently sparse, unordered and irregular, which limits the direct use of MLP-Mixer for point cloud understanding. To overcome these limitations, we propose PointMixer, a universal point set operator that facilitates information sharing among unstructured 3D point cloud. By simply replacing token-mixing MLPs with Softmax function, PointMixer can “mix" features within/between point sets. By doing so, PointMixer can be broadly used for intra-set, inter-set, and hierarchical-set mixing. We demonstrate that various channel-wise feature aggregation in numerous point sets is better than self-attention layers or dense token-wise interaction in a view of parameter efficiency and accuracy. Extensive experiments show the competitive or superior performance of PointMixer in semantic segmentation, classification, and reconstruction against Transformer-based methods.
Fast Point Transformer

Chunghyun Park, Yoonwoo Jeong, Minsu Cho, and Jaesik Park

IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2022
Silver Prize @ Samsung HumanTech Paper Award
Qualcomm Innovation Fellowship Winner
Abs arXiv Code Website

The recent success of neural networks enables a better interpretation of 3D point clouds, but processing a large-scale 3D scene remains a challenging problem. Most current approaches divide a large-scale scene into small regions and combine the local predictions together. However, this scheme inevitably involves additional stages for pre- and post-processing and may also degrade the final output due to predictions in a local perspective. This paper introduces Fast Point Transformer that consists of a new lightweight self-attention layer. Our approach encodes continuous 3D coordinates, and the voxel hashing-based architecture boosts computational efficiency. The proposed method is demonstrated with 3D semantic segmentation and 3D detection. The accuracy of our approach is competitive to the best voxel-based method, and our network achieves 129 times faster inference time than the state-of-the-art, Point Transformer, with a reasonable accuracy trade-off in 3D semantic segmentation on S3DIS dataset.