After days of searching I've found this video. Perfect, thank you! Question: "The number of (sub, mini) segments (e.g. one green MST) after the first step of the process, is equal to the number of seeds, the local minimas." Is this statement true? In other words: The number of seeds determines the level of under or over segmentation.
This lecture is a real gem, explains the intution behind a matlab aglorithem that most people know as a black box.
What is the subset at the time of 40:18?
Is there any threshold value as constraints to control the summary of weights from trees?
What are the background and foreground? Via which kind of operating approach can we differentiate them?
The cost of reaching node,to which node?From which node to which node?
Why do we use cost function here?We have already known to use minimum value to get the MST.
If we use prim's algorithm to expend the tree,but we don't have the threshold,what is the region of the tree?It could be all pixels.
@2:17, you may think of this as "ATLANTIS or whatever", hahahaha, you got great humour!!
If we use maximum path from the paper of Facoll ,whether can we use Kruskal's algorithm or not?
i want its code in opencv python
Great job 👍
After days of searching I've found this video. Perfect, thank you!
Question: "The number of (sub, mini) segments (e.g. one green MST) after the first step of the process, is equal to the number of seeds, the local minimas." Is this statement true?
In other words: The number of seeds determines the level of under or over segmentation.
How is the root node or meta node chosen? Based on what kind of method?