Shigon commited on
Commit
33149a0
·
verified ·
1 Parent(s): 9264105

Upload README.md with huggingface_hub

Browse files
Files changed (1) hide show
  1. README.md +41 -0
README.md ADDED
@@ -0,0 +1,41 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: other
3
+ task_categories:
4
+ - image-to-video
5
+ tags:
6
+ - 4dgrt
7
+ - 4d-gaussian-splatting
8
+ - ray-tracing
9
+ - camera-effects
10
+ - dynamic-scenes
11
+ - synthetic-data
12
+ - novel-view-synthesis
13
+ ---
14
+
15
+ # 4D-GRT Benchmark Data
16
+
17
+ This repository contains the synthetic benchmark data for the paper **Every Camera Effect, Every Time, All at Once: 4D Gaussian Ray Tracing for Physics-based Camera Effect Data Generation (4D-GRT)**.
18
+
19
+ [Project Page](https://shigon255.github.io/4DGRT-project-page/) | [arXiv](https://arxiv.org/abs/2509.10759)
20
+
21
+ 4D-GRT is a two-stage pipeline that combines 4D Gaussian Splatting with physically-based ray tracing for camera effect simulation. Given multi-view videos, it first reconstructs dynamic scenes, then applies ray tracing to generate videos with controllable, physically accurate camera effects. This repository provides the eight synthetic dynamic indoor scenes constructed as a benchmark to evaluate generated videos across four camera effects: pinhole, fisheye, depth of field, and rolling shutter.
22
+
23
+ ## Dataset Description
24
+
25
+ Each scene is rendered under 4 camera-effect variants (`pinhole`, `fisheye`, `dof`, `rolling_shutter`), each with 50 cameras and 50 frames at 512x512 resolution, along with a point cloud (`points3d_opencv.ply`) and `transforms_train.json` / `transforms_test.json` camera pose files. A shared `effect_params.csv` and `self_dataset_manifest.json` describe the per-scene effect parameters and manifest.
26
+
27
+ The 8 scenes are: `ball`, `basketball_1`, `basketball_2`, `box`, `cat`, `cube`, `lego`, `plant`.
28
+
29
+ ## Citation
30
+
31
+ ```bibtex
32
+ @misc{liu20254dgrt,
33
+ title={Every Camera Effect, Every Time, All at Once: 4D Gaussian Ray Tracing for Physics-based Camera Effect Data Generation},
34
+ author={Yi-Ruei Liu and You-Zhe Xie and Yu-Hsiang Hsu and I-Sheng Fang and Yu-Lun Liu and Jun-Cheng Chen},
35
+ year={2025},
36
+ eprint={2509.10759},
37
+ archivePrefix={arXiv},
38
+ primaryClass={cs.CV},
39
+ url={https://arxiv.org/abs/2509.10759},
40
+ }
41
+ ```