MYRIAD: Envisioning the Future, One Step at a Time

Stefan A. Baumann^*,1,2 · Jannik Wiese^*,1,2 · Tommaso Martorella^1,2 · Mahdi M. Kalayeh³ · Björn Ommer^1,2

¹CompVis @ LMU Munich, ²MCML, ³Netflix
CVPR 2026

From a single image MYRIAD predicts distributions over sparse point trajectories autoregressively. This allows us to predict physically consistent futures in open-set environments (top) conditioned on input movements. By exploring directly in motion space, we can rapidly explore thousands of counterfactual futures, enabling planning by search - here to select a billiard shot (bottom).

Note

This repository is a landing page for Myriad and primarily exists to host the project website.

➡️ The official code lives in CompVis/flow-poke-transformer.
Please refer to that repository for code, setup, and inference instructions.

🎓 Citation

@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}
}