What is the difference between proposal-based approach and proposal-free approach?

From here it says that

Techniques to solve instance segmentation can be roughly grouped into two categories: proposal-based methods and proposal-free methods. In proposal-based methods, a set of object proposals and their classes are first predicted, then foreground-background segmentation in each bounding box is performed. The proposal-free approaches exclude the step of proposal generation.

What is "proposal" in this context? Also, how to "first predict their classes"? There is not much explanation about this topic on the internet and I would appreciate it if someone could explain the differences.

Topic image-segmentation object-detection computer-vision deep-learning

Category Data Science

1
answered at

Proposal-based:

  1. Let's look for a car, define its boundary etc. Okay, found a car.
  2. Cluster all the pixels belonging to that car.

Proposal-free:

  1. Let's label each pixel as some uncategorized instance.
  2. Based on the results from semantic segmentation, that instance probably belongs to the "Car" category.

You can also check the II. Related Work in the paper below where it is explained in better detail, with some additional sources mentioned in it:

Hsu, Yen-Chang, et al. "Learning to cluster for proposal-free instance segmentation." 2018 International Joint Conference on Neural Networks (IJCNN). IEEE, 2018.

About

Geeks Mental is a community that publishes articles and tutorials about Web, Android, Data Science, new techniques and Linux security.