We developed the Single Slice Segmentation module for the task of segmenting time sequence of 2D frames, such as ultrasound image acquisitions.
Making it easy to copy the previous frame would introduce bias, because users would be tempted to accept the previous contour as is, even if it is slighly inaccurate. You would also end up with highly redundant data set, because neighbor frames in a time sequence are typically very similar. Therefore, we do not make it easy for users to copy the segment from the previous time point but instead we make it easy to skip frames (segment every N-th frame of the input sequence).