Using external datasets for pre-training models ¶
By: aneeqzia_isi on July 14, 2022, 5:14 a.m.
Hi everyone,
There were multiple teams who reached out regarding data allowed for pre-training models so we wanted to clarify that for everyone.
Teams are free to use any public dataset (including previous challenges) for pre-training their models. If any team has private data that they are pre-training their models on, then that data will need to be disclosed and publicly released upon submission to the challenge.
Hope this clarifies any confusion.
Best, SurgToolLoc 2022 Organizing Committee