remindof / ev-motionseg Goto Github PK

Hello, I'm reading the paper and I'm curious about the rotational motion model.

The paper says that the fan is represented by a rotational motion model and the coin by a linear velocity motion model, but doesn't give the details.
The paper A Unifying Contrast Maximization Framework for Event Cameras, with Applications to Motion, Depth, and Optical Flow Estimation gives the rotational model for camera ego-motion. I wonder whether this is the same.
Hope to hear your opinion, thanks a lot~

Hi, good job, but where can I find the official code, or it just not exists?

Hello, I wonder why this paper doesn't consider the polarity of events but manages to works well.

Hi, thanks for your code work. I just wonder what is the dataset in test-data? Maybe it is the hand sequence of DVSMOTION20 in DistSurf? Looking forward to your reply. Thank you in advance.

It seem that the function "updataMotionParam" won't run in the main loop, because you comment out this line in main.m file.
Besides, I am wondering what is your method to calculate this derivative of theta. In function "findGrad", it shows that you set a threshold for dist_row and dist_col and a triplet loop is set for the accumulation of derivate. But I cannot figure it out.
Could you give a more specific explanation for the method of this derivative calculate? Great appreciate!