TOPSPIN: daTaset Of sPinning objectS with neuromorPhic vIsioN

<p>Dataset used in the paper "<a href="https://arxiv.org/abs/2405.06216">Event-based Structure-from-Orbit</a>", for the purpose of sparse point cloud reconstruction of a spinning object.</p> <p>The dataset consists of six objects rotating at three speeds, being viewed by a static event camera from four different perspectives. For each scene, one of the objects was placed on the turntable and rotated at a given speed, while the camera observed it from a given perspective. We exhaustively enumerated all combinations of objects, rotational speeds and perspectives to generate 72 scenes in total.</p> <p>&nbsp;</p> <p>Scene names use the following encoding: {object}-{perspective}-{speed}</p> <p>All scenes have the same data format:</p> <p>scene/</p> <p>&nbsp;&nbsp;&nbsp;&nbsp;events.csv -- Event stream with the format {timestamp, x, y, polarity (0=off, 1=on)}</p> <p>&nbsp;&nbsp;&nbsp; calib.txt -- Camera calibration file, with the format "fx fy cx cy k1 k2 p1 p2 k3"</p> <p>&nbsp;&nbsp;&nbsp;&nbsp;meta.json -- Metadata file with camera frame dimensions</p> <p>Note: all timestamps are in microseconds.</p> <p>&nbsp;</p> <p><strong>When using the data in an academic context, please cite the following paper.</strong></p> <pre><code>@inproceedings{elms2024eventbased,<br>&nbsp; &nbsp; &nbsp; title={Event-based Structure-from-Orbit},&nbsp;<br> &nbsp; &nbsp;booktitle={IEEE Conf. on Computer Vision and Pattern Recpgnition},<br>&nbsp; &nbsp; &nbsp; author={Ethan Elms and Yasir Latif and Tae Ha Park and Tat-Jun Chin},<br>&nbsp; &nbsp; &nbsp; year={2024},<br>}</code></pre>