I learn a lot from many TD’s contribution on internet since the first day I step into CG industry, now it’s my turn to give something back to the internet.
When you’re reading this article,you may probably have watched this demo - Deformation Learning Solver, and you may curious how to make this tool.For a PhD majors in graphics, it’s no doubt that figuring out this algorithm in no time, if you’re a TD but not majors in math(like me), siggraph paper may cause you frustrate. So I decide to share my experiences to anyone who interesting in this technique, the following are MOOCs I’ve learned or I’ll learn.
Fundamental:
- Calculus One
- Calculus Two: Sequences and Series
- Coding the Matrix: Linear Algebra through Computer Science Applications
- Linear Algebra: Foundations to Frontiers
- Discrete Time Signals and Systems, Part 1: Time Domain
- Discrete Time Signals and Systems, Part 2: Frequency Domain
Intermediate:
- Machine Learning
- Machine Learning Foundations
- Machine Learning Techniques
- Linear and Integer Programming
- Discrete Optimization
- Digital Signal Processing
Advance:
Even you finish all courses above, it still can’t ensure you would understand SSDR, but they’ll get you the insight, the next thing is to prepare some books about Numerical Optimization and dig into the papers.
Actually, I didn’t implement SSDR at the very first beginning, since it’s still too hard for me, so I tried the earliest paper about example-based skinning which is call “Building Efficient, Accurate Character Skins from Examples”, this one is much easier, followed by “Skinning Mesh Animations”, then I can read “Skinning Arbitrary Deformations” and “Fast and Efficient Skinning of Animated Meshes”, finally, move straight to SSDR. It’s a tough time with no one’s help, if anyone who wants to implement this algorithm, I would love to share my experience, just contact me or leave comments.
Good luck!
近期评论