My topics of interest lie in efficient 3D representations,
with a particular focus on compression and acceleration of NeRF and Gaussian Splatting.
I'm always open to research collaborations — feel free to reach out anytime!
CoherentRaster is an efficient 3D Gaussian Splatting framework tailored for light field displays. By exploiting view coherence across the light field, our method renders high-quality multi-view content with significantly reduced computational cost.
Leveraging Learned Image Prior for 3D Gaussian Compression
We compress 3D Gaussian Splatting representations using a learned image prior. Our restoration model recovers high-frequency details lost during compression, enabling substantial storage reduction while preserving rendering quality.
Locality-aware Gaussian Compression for Fast and High-quality Rendering
We present a locality-aware compression scheme for 3D Gaussian Splatting. By leveraging the spatial locality of Gaussian primitives, LocoGS achieves compact storage and fast, high-quality rendering.
Binary Radiance Fields (BiRF) represent 3D scenes using binary feature grids. By quantizing features to binary values, BiRF achieves storage-efficient radiance fields while maintaining competitive rendering quality.
Deep 3D Reconstruction of Synchrotron X-ray Computed Tomography for Intact Lungs
Seungjoo Shin*, Min Woo Kim*, Kyong Hwan Jin, Kwang Moo Yi, Yoshiki Kohmura, Tetsuya Ishikawa, Jung Ho Je, Jaesik Park(*Equal contribution)
We propose a deep-learning-based 3D reconstruction approach for synchrotron X-ray Computed Tomography of intact lungs. Our method recovers fine anatomical structures from sparse projections, advancing high-resolution biological imaging.
PeRFception is a large-scale dataset of implicit 3D scene representations using radiance fields, designed to enable downstream perception tasks such as classification and segmentation directly on neural radiance fields.
Open-Source Library
NeRF-Factory: An awesome PyTorch NeRF collection
A library providing easily extensible and usable PyTorch-implementation of representative NeRF models.
