Datasets:
idx stringlengths 3 8 | answer stringlengths 0 19.7k |
|---|---|
0_i | C |
3_i | A |
4_i | B |
7_i | D |
8_i | A |
9_i | C |
10_i | B |
11_i | C |
12_i | A |
13_i | C |
14_i | A |
15_i | C |
16_i | B |
17_i | A |
19_i | A |
20_i | B |
26_i | A |
27_i | B |
28_i | C |
29_i | C |
30_i | B |
31_i | A |
32_i | C |
33_i | B |
34_i | D |
35_i | B |
43_i | A |
44_i | D |
49_i | C |
50_i | B |
51_i | C |
52_i | B |
53_i | C |
54_i | C |
55_i | D |
56_i | A |
57_i | B |
58_i | C |
59_i | C |
62_i | A |
63_i | D |
69_i | A |
70_i | B |
71_i | B |
72_i | A |
73_i | A |
74_i | A |
75_i | A |
76_i | A |
77_i | D |
78_i | D |
82_i | A |
83_i | A |
88_i | C |
89_i | B |
90_i | C |
91_i | B |
92_i | A |
93_i | B |
94_i | D |
95_i | A |
96_i | B |
97_i | B |
98_i | A |
99_i | A |
100_i | D |
106_i | C |
107_i | A |
108_i | A |
109_i | A |
110_i | C |
111_i | A |
112_i | C |
113_i | A |
114_i | D |
115_i | B |
116_i | D |
120_i | A |
125_i | A |
126_i | D |
131_i | A |
132_i | A |
133_i | D |
134_i | A |
135_i | A |
136_i | A |
137_i | D |
138_i | D |
139_i | C |
140_i | C |
141_i | C |
147_i | A |
148_i | D |
152_i | B |
153_i | B |
154_i | A |
155_i | D |
156_i | D |
157_i | A |
158_i | A |
NewtPhys dataset
This is the dataset used it NewtPhys for benchmarking VLMs ability to sense Newtonian physics. Its primary use is benchmarking and improving physical-awareness in vision models and generative AI.
The dataset provides 7k simulation sequences obtained by augmenting real-world scenes from DL3DV with Google Scanned Objects.
Each sequence comes with dense, low-level pixel-aligned ground truth maps — gravity, collisions, material segmentation, scene flow, optical flow, deformation gradients, instances, amodal maps, etc. — along with high-level annotations about physics — object interactions, velocity, names, etc. —.
The simulation code will be released soon too, along with added videos.
📄 Citation
If you use this codebase or the NewtPhys dataset, please cite:
@inproceedings{cavada2026newtphys,
title={{NewtPhys}: Do Foundation Models Understand Newtonian Physics?},
author={Sebastian Cavada and Soumava Paul and Tuan-Hung Vu and Andrei Bursuc and Raoul de Charette},
year=2026,
booktitle={arXiv},
url={https://arxiv.org/abs/2606.03986}
}
- Downloads last month
- 699