Hacker News new | ask | show | jobs
by throwaway84742 3014 days ago
Magic results but R-CNN (and other similar architectures with large number of forward passes per frame) is pretty darn slow, which severely limits its applicability.
2 comments

It really depends on the application. If you’re comparing to classical instance segmentation methods like a distance-based watershed, Mask RCNN (e.g. using the Keras-RCNN package) has nearly identical performance. However, if you’re looking for real-time performance, you’ll need to look elsewhere (and likely reframe your problem as an object detection rather than instance segmentation problem).
R-CNN is slow but Fast R-CNN and Faster R-CNN simply use one forward pass per image, making it feasible for use in real time