Benchmarking

This page aims to provide a centralized space for validation surgical vision algorithms against a common benchmark. We support user uploaded datasets (with associated citations for use) against which code can be tested. Leaderboards are available for each dataset showing the current state of the art against this dataset. If you would like to see any particular kind of surgical vision technique added to this benchmarking site, please feel free to send an email to anyone in our group with your request.

Our selection of datasets


Tracking data


2 datasets so far!

Detection data


6 datasets so far!