Meet your original competition.
Archive trajectories. Your move.
Loading the recordings…
Fetching the published clip list and researcher annotations.
How this works
Recorded video
A microscopy clip from a published research archive plays in your browser.
Published annotations
For each native frame, researcher-annotated boxes give each tracked cell a centre in image pixels.
Paddle movement
The selected cell’s normalised vertical position becomes the opponent paddle’s centre. Nothing else moves it.
Recorded research footage. Cell movement is replayed; the cell does not see or learn the game.
Science FAQ
Is the footage AI-generated?
No. The video is unmodified microscopy and the boxes are researcher-annotated trajectories. Nothing here is synthesised.
Do the cells learn to play Pong?
No. The cell cannot see or react to the game. Its recorded position simply moves a paddle.
Why can my opponent disappear?
Because the cell leaves the camera view or stops being annotated. When that happens the round freezes instead of inventing a position.
What do the coordinates measure?
Image pixels in the original video frame. Speed is shown in pixels per second only. There is no micron calibration.
Where is the data from?
The VISEM-Tracking dataset, published under CC BY 4.0. Files and annotations were adapted for browser playback.
Sources
VISEM-Tracking by Vajira Thambawita et al., SimulaMet and OsloMet. Scientific Data 10, 260 (2023).
- Paper: https://doi.org/10.1038/s41597-023-02173-4
- Dataset: https://zenodo.org/records/7293726
- Upstream code: https://github.com/simulamet-host/visem-tracking
- Data license: https://creativecommons.org/licenses/by/4.0/
Video files and annotations were adapted for browser playback. Data is used under CC BY 4.0; the original authors do not endorse this page.
The movement is biological.The game is ours.
No medical, diagnostic or fertility conclusions are offered here. No wallets, payments or wagering are involved.