Run on own data or Kitti

[russellyq]

Hi thanks for your great works ! @winstywang @ziqipang

I am trying to implement on my own dataset and kitti dataset. The detection results output "X, Y, Z, H, W, L, theta" and so does KITTI datasets.

As I see in your code that MOTModel updates the detection and return tracking results. https://github.com/TuSimple/SimpleTrack/blob/main/mot_3d/mot.py

How could I implement on own data with only detection output "X, Y, Z, H, W, L, theta" ?

Also, how to understand the use of ego motion in the code ?

Thanks.

[wenyiwangbst]

Hi, by any chance you are able to extract the labels from the .bin label file with the gt_bin_decode.py code?

[ziqipang]

@russellyq Hi, thanks for asking.

1. You are correct on using the API of MOTModel.frame_mot to run on each frame. The input to this function is FrameData. To get a better sense of what these things are, I think our demo is helpful.
2. In our code, ego motion refers to the pose or the location of the ego-vehicle in the global coordinate. It is generally represented by a 4-by-4 matrix. It is provided in all the benchmarks and is widely used in multi-frame LiDAR detection, such as CenterPoint.

[russellyq]

@ziqipang Thanks for your reply.

Does that mean I can use the ego motion from other benchmark as Waymo for my own detection result which only has 'X,Y,Z,theta,H,L,W' ?

[ziqipang]

@russellyq Sorry, I haven't understood your question. Maybe you can elaborate a little more, like what do you want/don't want ego-motion for?

[ziqipang]

Close due to inactivity.

[zjwzcnjsy]

你好，我已经接收到你的邮件！

[12w2]

Hi thanks for your great works ! @winstywang @ziqipang

I am trying to implement on my own dataset and kitti dataset. The detection results output "X, Y, Z, H, W, L, theta" and so does KITTI datasets.

As I see in your code that MOTModel updates the detection and return tracking results. https://github.com/TuSimple/SimpleTrack/blob/main/mot_3d/mot.py

How could I implement on own data with only detection output "X, Y, Z, H, W, L, theta" ?

Also, how to understand the use of ego motion in the code ?

Thanks.

hello. Do you replicate it on your own dataset

[mx2013713828]

@russellyq Hello, did you work it using kitti format result?