no code implementations • 12 Apr 2024 • Yuqun Wu, Jae Yong Lee, Chuhang Zou, Shenlong Wang, Derek Hoiem
Our experiments show 4x the performance of RegNeRF and 8x that of FreeNeRF on average F1@2cm for ETH3D MVS benchmark, suggesting a fruitful research direction to improve the geometric accuracy of NeRF-based models, and sheds light on a potential future approach to enable NeRF-based optimization to eventually outperform traditional MVS.
no code implementations • 13 Feb 2024 • Sung Woong Cho, Jae Yong Lee, Hyung Ju Hwang
There has been growing interest in models that learn the operator from the parameters of a partial differential equation (PDE) to the corresponding solutions.
no code implementations • 4 Feb 2024 • Michal Shlapentokh-Rothman, Ansel Blume, Yao Xiao, Yuqun Wu, Sethuraman T V, Heyi Tao, Jae Yong Lee, Wilfredo Torres, Yu-Xiong Wang, Derek Hoiem
We investigate whether region-based representations are effective for recognition.
no code implementations • 26 Dec 2023 • Jae Yong Lee, Sung Woong Cho, Hyung Ju Hwang
This study proposes HyperDeepONet, which uses the expressive power of the hypernetwork to enable the learning of a complex operator with a smaller set of parameters.
no code implementations • 9 Aug 2023 • Jae Yong Lee, Seungchan Ko, Youngjoon Hong
Partial differential equations (PDEs) underlie our understanding and prediction of natural phenomena across numerous fields, including physics, engineering, and finance.
no code implementations • 2 Dec 2022 • Jae Yong Lee, Yuqun Wu, Chuhang Zou, Shenlong Wang, Derek Hoiem
Instead, we propose to encode features in bins of Fourier features that are commonly used for positional encoding.
no code implementations • 2 Dec 2022 • Yuqun Wu, Jae Yong Lee, Derek Hoiem
Our long term goal is to use image-based depth completion to quickly create 3D models from sparse point clouds, e. g. from SfM or SLAM.
no code implementations • 14 Oct 2022 • Jae Yong Lee, Chuhang Zou, Derek Hoiem
Recent work in multi-view stereo (MVS) combines learnable photometric scores and regularization with PatchMatch-based optimization to achieve robust pixelwise estimates of depth, normals, and visibility.
no code implementations • 5 Jul 2022 • Jae Yong Lee, Juhi Jang, Hyung Ju Hwang
We propose a hybrid framework opPINN: physics-informed neural network (PINN) with operator learning for approximating the solution to the Fokker-Planck-Landau (FPL) equation.
1 code implementation • 28 Jan 2022 • Jin Young Shin, Jae Yong Lee, Hyung Ju Hwang
We combine the PDIO with the neural operator to develop a \textit{pseudo-differential neural operator} (PDNO) and learn the nonlinear solution operator of PDEs.
2 code implementations • CVPR 2022 • Liwen Wu, Jae Yong Lee, Anand Bhattad, YuXiong Wang, David Forsyth
DIVeR's representation is a voxel based field of features.
no code implementations • 9 Nov 2021 • Rakhoon Hwang, Jae Yong Lee, Jin Young Shin, Hyung Ju Hwang
Once the surrogate model is trained in Phase 1, the optimal control can be inferred in Phase 2 without intensive computations.
1 code implementation • ICCV 2021 • Jae Yong Lee, Joseph DeGol, Chuhang Zou, Derek Hoiem
To overcome the challenge of the non-differentiable PatchMatch optimization that involves iterative sampling and hard decisions, we use reinforcement learning to minimize expected photometric cost and maximize likelihood of ground truth depth and normals.
1 code implementation • CVPR 2021 • Jae Yong Lee, Joseph DeGol, Victor Fragoso, Sudipta N. Sinha
We address estimating dense correspondences between two images depicting different but semantically related scenes.
no code implementations • 28 Sep 2020 • Jae Yong Lee, Jin Woo Jang, Hyung Ju Hwang
The model reduction of a mesoscopic kinetic dynamics to a macroscopic continuum dynamics has been one of the fundamental questions in mathematical physics since Hilbert's time.
no code implementations • 22 Nov 2019 • Hyung Ju Hwang, Jin Woo Jang, Hyeontae Jo, Jae Yong Lee
The issue of the relaxation to equilibrium has been at the core of the kinetic theory of rarefied gas dynamics.