How do advances in machine learning or AI affect the use of kinematic animation, and are there any potential applications of these technologies in this field?
kkoujah
Advances in machine learning and AI have had a significant impact on kinematic animation, and there are many potential applications of these technologies in this field. One such application is in motion capture, where machine learning algorithms can be used to analyze motion capture data and extract meaningful information from it. This can help animators create more realistic and natural movements for characters and objects in their animations.
Spectato54
Inverse kinematics is sometimes used for rag doll effects in video games since it allows for a "realistic" animation. However, the inconsistency in physics during the animation allows for models to clip or phase through other models until the animation is done.
Noppapon
In order for kinematic animation to be consistent with physics, this can also add more work for artists, as they need to create complex simulations and interactions to make the animation look realistic and believable. Additionally, I've read how kinematic animation often requires a lot of trial and error to get the desired result. Animators may need to make many small adjustments to each joint and bone in order to achieve the desired movement or pose
modatberkeley
I agree, I think it's difficult to make kinematic animation look realistic. Is this the reason why we still see motion capture being used vs animating people/animals manually?
aceschen
A fun example of inverse kinematics in action is the loading screen of Super Mario 64, where you can stretch Mario's face! It ascribes to these pros and cons pretty clearly - it's so novel and iconic because it let's the player mess around pretty wildly, it doesn't need to really adhere to standard game physics, and it was probably a pain to make. More info about it here: https://www.nintendolife.com/news/2021/05/the_guy_who_made_the_stretchy_super_mario_64_face_also_almost_gave_us_portal_zelda
How do advances in machine learning or AI affect the use of kinematic animation, and are there any potential applications of these technologies in this field?
Advances in machine learning and AI have had a significant impact on kinematic animation, and there are many potential applications of these technologies in this field. One such application is in motion capture, where machine learning algorithms can be used to analyze motion capture data and extract meaningful information from it. This can help animators create more realistic and natural movements for characters and objects in their animations.
Inverse kinematics is sometimes used for rag doll effects in video games since it allows for a "realistic" animation. However, the inconsistency in physics during the animation allows for models to clip or phase through other models until the animation is done.
In order for kinematic animation to be consistent with physics, this can also add more work for artists, as they need to create complex simulations and interactions to make the animation look realistic and believable. Additionally, I've read how kinematic animation often requires a lot of trial and error to get the desired result. Animators may need to make many small adjustments to each joint and bone in order to achieve the desired movement or pose
I agree, I think it's difficult to make kinematic animation look realistic. Is this the reason why we still see motion capture being used vs animating people/animals manually?
A fun example of inverse kinematics in action is the loading screen of Super Mario 64, where you can stretch Mario's face! It ascribes to these pros and cons pretty clearly - it's so novel and iconic because it let's the player mess around pretty wildly, it doesn't need to really adhere to standard game physics, and it was probably a pain to make. More info about it here: https://www.nintendolife.com/news/2021/05/the_guy_who_made_the_stretchy_super_mario_64_face_also_almost_gave_us_portal_zelda