The challenge aims to explore approaches to action detection that leverage information across domains in a multi-task setting.
Based on two datasets of videos of radical prostatectomy procedures captured on both real and artificial data, the challenge will provide valuable insight into how to learn surgical action detection models robust to domain variations.

The most novel and the top-performing methods will be invited to deliver a talk at the Challenge event.
Their authors will be invited as co-authors in a journal paper dedicated to the Challenge.

Detailed instructions on the tasks, how to download the data and submit your predictions for evaluation are provided on the website.

Dates

Challenge opens for registration: April 1, 2021
Training/Validation data release: April 27, 2021
Test data release: July 16, 2021
Result submission deadline: July 31, 2021
Final result announcement: August 5, 2021
Challenge date: September 27 – October 1, 2021