[question] featureExtraction cache

[jonnytest1]

Describe the problem
so ive run the whole photogrammetry pipeline for a set of images and then when i add some more images i would expect the feature extraction step (which i imagine doesnt havve any "cross-image" matching yet) to only run on the newly added images 🤔 (im not currently using the live restronstruction mode tho because that doesnt generate a mesh at the end)