PGET challenge results announced
After evaluation of all submitted solutions against comprehensive criteria (quality of reconstructed image, i.e. contrast, spatial resolution, mitigation of artifacts; accuracy of processing, i.e. pin counts, separation of missing and anomaly pins, activity reconstruction; computational speed, complexity of setup, need for a-priori information; innovative content of the approach), the top 3 ranked solutions are selected. The laureates of the IAEA PGET challenge are:
- Dr Luca Presotto, Nuclear Medicine Unit of IRCCS Ospedale San Raffaele, Italy
- Rasmus Backholm, Helsinki Institute of Physics, University of Helsinki, Finland
- Dr Christian Thomay, Austria
The IAEA challenge team will invite the laureates for a technical meeting on PGET data processing to take place at the IAEA HQ (Vienna, Austria) in September 2019. Solutions to the competition datasets are now also available for all participants to analyze their algorithms and approaches.
IAEA Data Visualization Challenge launched
You have a passion for data visualization? This announcement is for you. Develop a compelling infographic, or a rich interactive data visualization, and win a trip to Vienna, Austria, to present it at the International Conference on Climate Change and the Role of Nuclear Power (7-11 October 2019). See more for details.
Tomography reconstruction and analysis challenge - deadline extended
Following the participation survey results, the IAEA challenge team has decided to extend the training phase of the Tomography reconstruction and analysis challenge to allow more time for developing and improving your solutions. We will be accepting your training solutions until 31 May 2019; and the competition phase will take place on 3 - 9 June 2019.
We invite you to take advantage of this extension and continue submitting your solutions for the training datasets – our benchmark is available online and is regularly updated. Please note that benchmarking is fully anonymous and does not count for the evaluation of the challenge – its sole goal is to help you fine-tune your algorithms before the competition starts.
Tomography reconstruction and analysis challenge - participation survey
Please take part in a short survey to improve the PGET challenge and ensure everyone can participate. The survey will take less than 2 minutes to complete. We thank you in advance for your participation!
Update on the IAEA Tomography reconstruction and analysis challenge
- The training phase will be concluded, and the competition will commence on 29 April 2019 (at noon CEST). New datasets will be made available to all participants in the "Downloads" section on the challenge details page. We will be accepting your solution until 6 May 2019 COB CEST at the submission page. We will send out reminders a few days before these milestones.
- In the meantime, we invite you to take advantage of our online benchmark tool (at the bottom of the challenge details page): submit at any time your solutions for the training datasets, and see how well they perform against IAEA's own algorithm and other submissions - this may help you fine-tune your algorithms before the competition starts. More importantly, this will ensure that your submission is formatted in accordance with the output data format requirements - improperly formatted data will be corrected with the participants during the training phase, but may be discarded during the competition. Please note that benchmarking is fully anonymous and does not count for the evaluation of the challenge.
- Don't forget to look at the updated FAQ page, where various submitted questions are addressed. Still have questions? Please post them using the online form at the bottom of the page.
- Finally, we have prepared a short survey to improve this challenge and ensure everyone can participate. The survey will take less than 2 minutes to complete. We thank you in advance for your participation!
Article on the results of the IAEA Robotics Challenge
PGET challenge is now open!
IAEA Robotics Challenge - winner announcement
After undergoing thorough evaluation of design and performance by IAEA experts, the winner of the IAEA Robotics Challenge is the Unmanned Surface Vehicle (USV) produced by Datastart Ltd of Hungary.
During Demonstration Week in Nov 2017, Datastart’s USV was selected alongside two other designs to continue in the competition. These designs then underwent a series of improvements and refinements based on the requirements of nuclear facility operators and IAEA suggestions. At the final round of the competition in November 2018, the designs underwent real-world testing in a spent fuel storage pond at a nuclear power plant in Finland. Datastart’s USV showcased its potential and impressed the IAEA’s team of experts. Our congratulations to Datastart Ltd!