Contribution 7: Filtering Methods

This will be my last post for the first phase of my GSoC journey. I had promised to deliver a working module that returns the count of people from a single view through a continuous video feed. From my earlier post, you might have seen the challenges I had faced and the solutions I had proposed. I had mentioned that use of filtering techniques may help smooth out unrealistic jumps in a continuous video feed, I had resorted to two major methods which proved to improve the performance of the count values return with the SS-DCNet model namely,

1. 1D Kalman Filter based approach
2. A moving average based approach

Both these methods can be used depending on the crowd density environment and in some cases, the count value returned by the SS-DCNet is itself pretty much good. I have analyzed both these methods with SALSA dataset which are shown below:

I have also updated the code which now takes in an extra argument which is the filtering method.

I am truly happy as the first phase of the project comes to a close and I really hope I have been able to do justice to it. Looking forward to the next phase!! 😃

Status: Reviewed