From what I can see is that in the cost functions, there are two versions of which only collision cost is using a kernel smoothening function. Why is no kernel smoothening function implemented for obstacle distance gradient? Would this not improve the performance when using this cost function?
From what I can see is that in the cost functions, there are two versions of which only collision cost is using a kernel smoothening function. Why is no kernel smoothening function implemented for obstacle distance gradient? Would this not improve the performance when using this cost function?