Is there any constraint on the use of temporal information? Are we allowed to use previous or future frames during prediction? Also, do the architectures have to be the same for both categories or can they be different?

Created by Raabid Hussain raabid236
Hi @raabid236, Thanks for posting your questions. There is no constraint in the use of temporal information - you can use future or previous frames to make current frame predictions, as long as the final output for each video consists of predictions for all frames. The architectures do not need to be the same for both categories. However, please ensure that your methodology report reflects this and your docker container contains all the necessary files to run. Please feel free to reach out if you have any other questions or concerns. Best, SurgVisDom 2020 Organizing Committee

Query about use of temporal information page is loading…