With an impressive accuracy of 91.15%, the team winning the competition for top performance is COSBI composed by Valerio Guarrasi and Paolo Soda from Italy Campus Biomedico, Congratulations! However, due to their previous experience on the data and previous work on the AIforCOVID dataset, the members of COSBI decided to renounce to the prize. In any case, they provide the software openly for repeatability purposes (code, models and XAI, report).
Thus the Prize of 5.000,00€ kindly offered by Bracco Imaging for best result goes to team csabAIbio from Hungary. Congratulations to Alex Olar, András Biricz, Bendegúz Sulyok, and Zsolt Bedőházi! They also provide their software openly (code, checkpoints, report).
The complete leaderboard is available here.
Results in terms of Explainability, Data Interpretability has been evaluated by a teams of both Computer Scientists, Radiologists and Health experts*.
The three best explainability results are given by the following teams:
According with the experts in the jury, the prize of 2.500,00€ for “accuracy within a human-centric AI solution” sponsored by CINI Lab AIIS goes to CoppolaFerrari from Italy. Congratulations to Coppola Edoardo and Ferrari Damiano!
The complete submission list is available here.
Tessadori Jacopo, University of Verona
Diego Sona, Fondazione Bruno Kessler
Rita Cucchiara, University of Modena and Reggio Emilia
Alessio Del Bue, Italian Institute of Technology
Costantino Grana, University of Modena and Reggio Emilia
Fazzini Deborah, Centro Diagnostico Italiano
Michaela Cellina, ASST Fatebenefratelli Sacco
Ali Marco, Centro Diagnostico Italiano
Valbusa Giovanni, Bracco imaging