README.md

---
language:
- en
license: cc-by-nc-sa-4.0
size_categories:
- n<1K
task_categories:
- other
pretty_name: MYRIAD-Physics
---

# MYRIAD-Physics


[![Project Page](https://img.shields.io/badge/Project-Page-blue)](https://compvis.github.io/myriad)
[![Paper](https://img.shields.io/badge/arXiv-paper-b31b1b)](https://arxiv.org/abs/2604.09527)
[![Paper](https://img.shields.io/badge/Huggingface-Papers-yellow)](https://huggingface.co/papers/2604.09527)
[![GitHub](https://img.shields.io/badge/GitHub-Code-black)](https://github.com/CompVis/flow-poke-transformer)
[![MYRIAD Weights](https://img.shields.io/badge/HuggingFace-Weights-orange)](https://huggingface.co/CompVis/myriad)
[![OWM Benchmark](https://img.shields.io/badge/Related-OWM--95-green)](https://huggingface.co/datasets/CompVis/owm-95)


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).

## Abstract

MYRIAD-Physics extends Physics-IQ and Physion with motion annotations and object tracks (following the same approach proposed in [`CompVis/owm-95`](https://huggingface.co/datasets/CompVis/owm-95)) for evaluating probabilistic future trajectory forecasting under physical interactions.

Unlike [`CompVis/owm-95`](https://huggingface.co/datasets/CompVis/owm-95), which distributes videos together with annotations, this repository provides only the additional metadata: annotations and trajectories for videos that must be obtained separately using [this download script](https://github.com/CompVis/flow-poke-transformer/blob/main/scripts/myriad_eval/download_datasets.sh).

We manually annotate relevant objects and the type of motion observed, and we use an off-the-shelf tracker to obtain motion trajectories with manual verification of correctness.

## Project Page and Code

- Project Page: https://compvis.github.io/myriad
- GitHub Repository: https://github.com/CompVis/flow-poke-transformer

## Usage

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).

To run the benchmark evaluation for this dataset, you can use the following command:

```shell
python -m scripts.myriad_eval.openset_prediction --data-root path/to/data  --ckpt-path path/to/checkpoint --dataset-name [owm | physion | physics-iq]
```

## Citation

If you find our data or code useful, please cite our paper:

```bibtex
@inproceedings{baumann2026envisioning,
  title={Envisioning the Future, One Step at a Time},
  author={Baumann, Stefan Andreas and Wiese, Jannik and Martorella, Tommaso and Kalayeh, Mahdi M. and Ommer, Bjorn},
  booktitle={Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
  year={2026}
}
```