Graph Matching

List broad steps until MICCAI

Active Graph Matching

Unary Costs defined on Location, Shape ... Binary Costs. All weighted by lambdas, which are set.

Data: 30 worms, C. elegans L1 larvae.

Questions:

1. Access to the actual images of 30 worms?

2. Access to the DD implementation? Is AMIRA plugin opensource?

Idea:

1. Use StarDist3D and deep graph matching for joint segmentation and annotation!

Learning Graph Matching, 2009

"We argue that, if we know the “conditions” under which a pair of graphs have been extracted, then we should take into account how graphs arising in those conditions are typically matched."

"A major finding in this work has been the realization that linear assignment with learning may perform similarly or better than state-of-the-art quadratic assignment relaxation algorithms (without learning)."

Questions:

Is it interesting to show that shallow learning benefits on the C. elegans dataset? Any other 3D datasets? Semi-supervised learning on Platynereis?