Changil Kim

[Changil Kim @_ChangilKim]

I am hiring research interns for 2025 at Meta. Please apply via the following link and drop me a line if you're interested in working on cutting-edge research problems in neural rendering, 3D reconstruction, gen AI, etc.
https://t.co/Yxcz7CvOPo

[Jia-Bin Huang @jbhuang0604]

3D Gaussian Splatting is awesome! BUT, the 3D Gaussian primitives often have difficulty reconstructing fine appearance details. 😩

Meet Textured Gaussians! ✋

Our RGBA Textured Gaussians enhance 3D scene appearance modeling. Small change, BIG improvement!

[Changil Kim @_ChangilKim]

Check out our work on Textured Gaussians that I've been cooking with my brilliant intern @BrianCChao and other amazing collaborators!

Paper: https://t.co/xyUbGxHnjr
Webpage: https://t.co/iEUEOvqRgD

[Brian Chao @BrianCChao]

Excited to share our new paper on improving Gaussian splatting novel-view synthesis quality through augmenting Gaussians with RGBA texture maps! (1/n)

[Jia-Bin Huang @jbhuang0604]

Beyond photos/videos, what's the next medium for sharing visual experiences? 🤔

NeRF/3DGS are great but only capture a static 3D world and ignore ambient scene dynamics.

In this paper, we show how we can reconstruct/render these essential scene elements even with a single causal video.

[Changil Kim @_ChangilKim]

@wenzeljakob Congrats, Wenzel!

[Changil Kim @_ChangilKim]

We're looking for summer interns 2024 in Reality Labs at Meta. Apply via the link below and reach out to me if you're excited about neural rendering, scene reconstruction, view synthesis, and generative models.
https://t.co/5HGzzz1KME

[Changil Kim @_ChangilKim]

Check out our work on real-time NeRF rendering for VR!

https://t.co/SxreTyoOah

[Christian Richardt @c_richardt]

Introducing VR-NeRF (@SIGGRAPHAsia 2023):
🤩 multi-camera rig for multi-view HDR capture
🤓 perceptual HDR optimization + level of detail
😎 real-time multi-GPU VR rendering

Project: https://t.co/2Kjpez8mcK
Paper: https://t.co/5ZSQ0vp1qv
Dataset: https://t.co/nngHuWpRjX

[Changil Kim @_ChangilKim]

Please check our latest work!

[Jia-Bin Huang @jbhuang0604]

Introducing OmnimatteRF!

From a coarse mask as input, our method produces detailed video matting (RGB alpha) of the subject *and* their associate effects, e.g., soft cast shadows.

This enables a variety of cool video editing applications.

[Changil Kim @_ChangilKim]

Time flies!

[International Conference on 3D Vision (3DV) @3DVconf]

Best Paper Award at #3DV2016

Depth from Gradients in Dense Light Fields for Object Reconstruction

Kaan Yucer, Changil Kim @_ChangilKim, Alexander Sorkine-Hornung, Olga Sorkine-Hornung @OlgaSorkineH

#TBThursday #3DV2024

[Changil Kim @_ChangilKim]

Andreas is presenting our work on Local Radiance Fields at CVPR this morning! Come to Poster 6 to check our work!

Project page: https://t.co/CGsf43AYFB

[Jia-Bin Huang @jbhuang0604]

Turn your casual videos into immersive 3D rendering!

How?
👉 Modeling scenes with multiple LOCAL radiance fields
👉 Optimizing poses PROGRESSIVELY

See you at AM poster session today! 👋

Video 📼: https://t.co/Si1yj3fBVz
Web 🔗: https://t.co/4k5bPZxlKs

#CVPR2023

[Changil Kim @_ChangilKim]

Ben (@imarhombus) is presenting our CVPR highlight paper about HyperReel, a high-fidelity 6-DoF neural volumetric video this morning in the exhibition hall (Poster 13). Come check out our work!

Project webpage: https://t.co/vIaORpRe4B

[Jia-Bin Huang @jbhuang0604]

🔥 New video 🔥

Check out HyperReel: High-Fidelity 6-DoF Video!
(#CVPR2023 highlight)

HyperReel achieves
✅ Fast rendering
✅ High-fidelity free-view synthesis
✅ Memory efficiency

https://t.co/oK3BxOtj2w

[Changil Kim @_ChangilKim]

We presented our work on joint optimization of spatiotemporal RF and camera poses of dynamic scenes on Tuesday at CVPR.

The code is available now! See the project webpage: https://t.co/0O1SPGU0bk

[Jia-Bin Huang @jbhuang0604]

SUPER excited to present ROBUST Dynamic Radiance Fields!

Existing methods cannot handle casual videos where SfM does not work.

Our work improves the robustness and can create 3D videos from ANY video.

Come find us in the AM session! 👋

🔗 https://t.co/AQW3fmPxU4

#CVPR2023

[Jon Barron @jon_barron]

#CVPR2023 is mostly posters this year, which is good! Poster sessions can be great. But a lot of people are not so great at presenting their poster. Here's some of the advice I give grad students when preparing for poster sessions.

[Jia-Bin Huang @jbhuang0604]

We just released the code, pretrained models (with MIT license 🔥)
💻 https://t.co/z0YnzEZOx3

as well as all the video data/results used in our work
📁 https://t.co/KeLjy2T0Db

Check out the explainer video
📹 https://t.co/cNdSB4Eutx

[Jia-Bin Huang @jbhuang0604]

🎉 New video 🎉

Check out the video to learn how to create immersive 3D rendering from casual videos! 🤩

https://t.co/Si1yj3fBVz

[Jia-Bin Huang @jbhuang0604]

Check out ✨HyperReel✨ (#CVPR2023 highlight)!

HyperReel is a 6-DoF video representation with
🤩 high-fidelity quality,
😍 fast rendering speed, and
😎 small memory footprint!

Project: https://t.co/mKbAUcIULW
Paper: https://t.co/RW7VevMFlV
Code: https://t.co/N9xjvdITVR

[Jia-Bin Huang @jbhuang0604]

Check out our fun work on robust view synthesis! 🤩

Using a casual video as input, our method jointly 1⃣estimates accurate camera poses and
2⃣reconstructs multiple local radiance fields for a large scene.

Project🕸️: https://t.co/4k5bPZxlKs

[Jia-Bin Huang @jbhuang0604]

Excited to share our #CVPR2023 on synthesizing new views along a camera trajectory from a **single image**!

How?
💡 The good old epipolar constraints in a pose-guided diffusion model!

Paper: https://t.co/B5rGOHRfcE
Project: https://t.co/ududk1pDpv

[Changil Kim @_ChangilKim]

We've just released the code & all results for our Neural Light Fields work that will appear at CVPR 2022!

Code & Results: https://t.co/UUsYmcFYmL
Project: https://t.co/OzBehwHtgJ

@imarhombus @jbhuang0604 @MZollhoefer @JPKopf

[Changil Kim @_ChangilKim]

Praying for peace...

[Ankita Jain @Ankita20200]

Little girl singing "Let it go" in a shelter

#UkraineRussianWar #Ukraine #UkraineUnderAttack