Datasets:

Tasks:
Other
Languages:
English
Size:
n<1K
ArXiv:
License:
tommymarto nielsr HF Staff commited on
Commit
b8c8f93
·
1 Parent(s): f6c4a52

Add task category and sample usage (#1)

Browse files

- Add task category and sample usage (f3bf5b4b3e72fb25dc8c8ec874af4d794d8c0a65)


Co-authored-by: Niels Rogge <nielsr@users.noreply.huggingface.co>

Files changed (1) hide show
  1. README.md +11 -2
README.md CHANGED
@@ -1,10 +1,12 @@
1
  ---
2
- license: cc-by-nc-sa-4.0
3
  language:
4
  - en
5
- pretty_name: MYRIAD-Physics
6
  size_categories:
7
  - n<1K
 
 
 
8
  ---
9
 
10
  # MYRIAD-Physics
@@ -13,6 +15,7 @@ size_categories:
13
  [![Paper](https://img.shields.io/badge/Paper-CVPR%202026-red)](https://arxiv.org/abs/2604.09527)
14
  [![OWM Benchmark](https://img.shields.io/badge/Related-OWM--95-green)](https://huggingface.co/datasets/CompVis/owm-95)
15
 
 
16
 
17
  ## Abstract
18
 
@@ -31,6 +34,12 @@ We manually annotate relevant objects and the type of motion observed, and we us
31
 
32
  We provide code and instructions to download the dataset and run the MYRIAD evaluation in our [GitHub repository](https://github.com/CompVis/flow-poke-transformer).
33
 
 
 
 
 
 
 
34
  ## Citation
35
 
36
  If you find our data or code useful, please cite our paper:
 
1
  ---
 
2
  language:
3
  - en
4
+ license: cc-by-nc-sa-4.0
5
  size_categories:
6
  - n<1K
7
+ task_categories:
8
+ - other
9
+ pretty_name: MYRIAD-Physics
10
  ---
11
 
12
  # MYRIAD-Physics
 
15
  [![Paper](https://img.shields.io/badge/Paper-CVPR%202026-red)](https://arxiv.org/abs/2604.09527)
16
  [![OWM Benchmark](https://img.shields.io/badge/Related-OWM--95-green)](https://huggingface.co/datasets/CompVis/owm-95)
17
 
18
+ MYRIAD-Physics extends Physics-IQ and Physion with motion annotations and object tracks for evaluating probabilistic future trajectory forecasting under physical interactions. It was presented in the paper [Envisioning the Future, One Step at a Time](https://huggingface.co/papers/2604.09527).
19
 
20
  ## Abstract
21
 
 
34
 
35
  We provide code and instructions to download the dataset and run the MYRIAD evaluation in our [GitHub repository](https://github.com/CompVis/flow-poke-transformer).
36
 
37
+ To run the benchmark evaluation for this dataset, you can use the following command:
38
+
39
+ ```shell
40
+ python -m scripts.myriad_eval.openset_prediction --data-root path/to/data --ckpt-path path/to/checkpoint --dataset-name [owm | physion | physics-iq]
41
+ ```
42
+
43
  ## Citation
44
 
45
  If you find our data or code useful, please cite our paper: