I'm curious as to why the hierarchical representation of a skeleton helps with forward and inverse kinematics calculations. Does indexing // structuring the model help similar in the way a BVH allows efficient finding of certain parts of the body? Or does applying transformations work by applying transformations to each group? Then, what order should the transformations be executed in so that the resulting effect is accurate.
S-Muddana
The hierarchical representation of a skeleton aids both forward and inverse kinematics calculations by organizing joints into a structured hierarchy, akin to a BVH in ray tracing. In forward kinematics, transformations are propagated from parent to child joints, while in inverse kinematics, they are adjusted iteratively from child to parent joints. This hierarchical approach ensures efficient traversal and manipulation of the skeleton's components, maintaining accurate spatial relationships between joints. Transformations are applied consistently and hierarchically to achieve precise positioning and movement of the skeleton.
s3kim2018
With hierarchical representation, I think it might be harder to model dynamic movements (not impossible). For example if you move your shoulder, you essentially move your whole arm. What if you want to move your arm outwards and push your elbows inwards at the same time? Coding this will not be intuitive and would have to account for the parent's movement.
aishikbhattacharyya
In animation software, mapping your skeleton exactly like how it's shown in the slide is good practice for organizing all your parts. A lot of the time, you're copying images from one half of the body to another so by keeping it organized like this, you can distinguish the left from the right very easily.
snowshoes7
In defense of this kind of construction I do think it semi decently represents the bulk of what actual movements are like. It's physically not so common to move a specific subpart of one of these differently from its parent, so in most cases I'd suspect this would make animation easier.
I'm curious as to why the hierarchical representation of a skeleton helps with forward and inverse kinematics calculations. Does indexing // structuring the model help similar in the way a BVH allows efficient finding of certain parts of the body? Or does applying transformations work by applying transformations to each group? Then, what order should the transformations be executed in so that the resulting effect is accurate.
The hierarchical representation of a skeleton aids both forward and inverse kinematics calculations by organizing joints into a structured hierarchy, akin to a BVH in ray tracing. In forward kinematics, transformations are propagated from parent to child joints, while in inverse kinematics, they are adjusted iteratively from child to parent joints. This hierarchical approach ensures efficient traversal and manipulation of the skeleton's components, maintaining accurate spatial relationships between joints. Transformations are applied consistently and hierarchically to achieve precise positioning and movement of the skeleton.
With hierarchical representation, I think it might be harder to model dynamic movements (not impossible). For example if you move your shoulder, you essentially move your whole arm. What if you want to move your arm outwards and push your elbows inwards at the same time? Coding this will not be intuitive and would have to account for the parent's movement.
In animation software, mapping your skeleton exactly like how it's shown in the slide is good practice for organizing all your parts. A lot of the time, you're copying images from one half of the body to another so by keeping it organized like this, you can distinguish the left from the right very easily.
In defense of this kind of construction I do think it semi decently represents the bulk of what actual movements are like. It's physically not so common to move a specific subpart of one of these differently from its parent, so in most cases I'd suspect this would make animation easier.