Sangeek Hyun

I am a Ph.D. candidate in the Visual Computing Lab (VCLab) at Sungkyunkwan University, supervised by Prof. Jae-Pil Heo. I received my Master's and Bachelor's degrees from Sungkyunkwan University. My research interests include various tasks in machine learning and computer vision, with a particular focus on generative models and video understanding.

Email  /  Google Scholar  /  Github  /  LinkedIn

profile photo

Research

Recently my interest has been in 3D generative models using Generative Adversarial Networks and gaussian splatting. I am also interested in various generation tasks using large-scale diffusion models.

GSGAN: Adversarial Learning for Hierarchical Generation of 3D Gaussian Splats
Sangeek Hyun, Jae-Pil Heo
NeurIPS, 2024
project page / code / arXiv

First 3D GANs utilize gaussian splatting without any structural priors, achieving faster rendering speed at high-resolution data compared to NeRFs.

Style Injection in Diffusion: A Training-free Approach for Adapting Large-scale Diffusion Models for Style Transfer
Jiwoo Chung*, Sangeek Hyun*, Jae-Pil Heo (*: Equal contribution)
CVPR, 2024 (Highlight)
project page / code / paper / arXiv

Training-free style transfer utilizes large-scale diffusion models by manipulating the attention features.

Diversity-aware Channel Pruning for StyleGAN Compression
Jiwoo Chung, Sangeek Hyun, Sang-Heon Shim, Jae-Pil Heo
CVPR, 2024
code / paper / arXiv

GAN compression technique by pruning the diversity-aware channels in StyleGAN architecture.

Task-disruptive Background Suppression for Few-Shot Segmentation
Suho Park, SuBeen Lee, Sangeek Hyun, Hyun Seok Seong, Jae-Pil Heo
AAAI, 2024
code / paper / arXiv

Task-disruptive Background Suppression mitigates the negative impact of dissimilar or target-similar support backgrounds, improving the accuracy of segmenting novel target objects in query images.

Correlation-guided Query-Dependency Calibration in Video Representation Learning for Temporal Grounding
WonJun Moon, Sangeek Hyun, SuBeen Lee, Jae-Pil Heo
Arxiv, 2023
code / arXiv

CG-DETR improves temporal grounding by using adaptive cross-attention and clip-word correlation to accurately identify video highlights corresponding to textual descriptions.

Frequency-based motion representation for video generative adversarial networks
Sangeek Hyun, Jaihyun Lew, Jiwoo Chung, Euiyeon Kim, Jae-Pil Heo
TIP, 2023
project page / code / paper

Propose a frequency-based motion representation for video GANs, enabling speed-aware motion generation, which improves video quality and editing capability.

Disentangled Representation Learning for Unsupervised Neural Quantization
Haechan Noh, Sangeek Hyun, Woojin Jeong, Hanshin Lim, Jae-Pil Heo
CVPR, 2023
paper

Disentangled representation learning for unsupervised neural quantization addresses deep learning quantizers' limitations in leveraging residual vector space, enhancing search efficiency and quality.

Query-dependent video representation for moment retrieval and highlight detection
WonJun Moon*, Sangeek Hyun*, SangUk Park, Dongchan Park, Jae-Pil Heo (*: Equal contribution)
CVPR, 2023
project page / code / paper / arXiv

Query-Dependent DETR improves video moment retrieval and highlight detection by enhancing query-video relevance and using negative pairs to refine saliency prediction.

Local attention pyramid for scene image generation
Sang-Heon Shim, Sangeek Hyun, DaeHyun Bae, Jae-Pil Heo
CVPR, 2022
paper

The Local Attention Pyramid (LAP) module addresses class-wise visual quality imbalance in GAN-generated scene images by enhancing attention to diverse object classes, particularly small and less frequent ones.

Self-Supervised Video GANs: Learning for Appearance Consistency and Motion Coherency
Sangeek Hyun, Jihwan Kim, Jae-Pil Heo
CVPR, 2021
paper

Self-supervised approaches with dual discriminators improve video GANs by ensuring appearance consistency and motion coherency through contrastive learning and temporal structure puzzles.

VarSR: Variational Super-Resolution Network for Very Low Resolution Images
Sangeek Hyun, Jae-Pil Heo
ECCV, 2020
paper

VarSR leverages latent distributions to address the many-to-one nature of single image super-resolution, generating diverse high-resolution images from low-resolution inputs.


Website template.