Use of previous frames?

In the rules section, under point 5 it is written:
“must not use future frames to infer for the current frame (usages of previous frames is allowed)”

This means that we can use the previous
frames in a flight to predict the depth of
current frame. However, it is not completely
clear how to use these frames, as in the
prediction method in the depth estimator it
only takes the current frame as input.

Can someone specify how we can use these
previous frames?

@nippun_sharma The original plan for the challenge was to provide video clips instead of images, but the problem statement was changed. I think the organizers forgot to remove that clause in the rules by mistake. Please ignore it.