Hi, thank you for sharing your code. I have test run the code on the full waymo validation dataset and using the detection you send to me(thanks for doing that) and use waymo evaluation metrics to do the evaluation. The result I got shows as following: But in the paper the L2 Vehicle could achieve around 0.56 in MOTA which is about 3.5% higher than my implement. Can I ask if the released code is of the same setting in parameter as the one in the paper and what evaluation method are you guys using. Thanks again for all your work and help.

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.

Hi, I use center point to generate output a bin output as well. I try to use the detection.py to export the detection information but return error during saving. I look at the code and seems the frame_number is always None. So ts_info[segment_name][_j] is never equal to current object's time_stamp. Can I ask how to solve this problem? Thank you so much for make the code public and for helping

Hi Ziqi,

I have one simple question regarding the NMS. Because the threshold_high is set to 1.0, it seems that the block of code below is never used in NMS: https://github.com/TuSimple/SimpleTrack/blob/3b44d6d197b06501f01b939e2f1da764e44ac5dd/mot_3d/preprocessing/nms.py#L47-L71 Could you please illustrate a bit more how and where the above code is useful?

Thanks!

Hi, in the paper the asso_thres is set as -0.5. However, I find the value is set as 1.5 in the config file. Is there any information I missing? asso_thres: iou: 0.9 giou: 1.5

Thanks for your wonderful work! Do you have the code for the constant velocity model on Nuscenes dataset? I tried to implement the constant velocity model but cannot reproduce the results. Is there any preprocessing of the velocity or time stamp? Thank you very much!

Hello, It seems that the data association is "one-stage" just using the high confidence threshold according to the configure file on nuscenes(the redundancy mod is default). Does it mean "one-stage" data association have best performance on nuscenes, rather than "two-stage" association which is used on waymo dataset? So Could you explain why, Is it related to frequency of Datasets? Thank you

Hi, First of all, thank you for sharing your code.

I have been trying to preprocess the WOD data but I am stuck on the second step (decode the ground truth), where I am getting the following error: Traceback (most recent call last): File "gt_bin_decode.py", line 129, in <module> main(args.file_path, out_folder, args.data_folder) File "gt_bin_decode.py", line 81, in main segment_name = segment_name_list[val_index] TypeError: list indices must be integers or slices, not NoneType

Any idea what could I be? I tried to track the error and I think it could have to do with the .json files, somehow.

nuScenes

1. Preprocessing

2. Detection python detection.py --raw_data_folder ${raw_data_dir} --data_folder ${data_dir_2hz} --det_name ${name} --file_path ${file_path} --mode 2hz --velo How can i get the detection file(in code, this default is val.json)

3. SimpleTrack/docs/nuScenes.md

python tools/nuscenes_result_creation.py \
    --name SimpleTrack2Hz \
    --name result_folder ${nuscenes_result_dir} \
    --data_folder ${nuscenes2hz_data_dir}

=================>>>>>>(i think)

python tools/nuscenes_result_creation.py \
    --name SimpleTrack2Hz \
    --result_folder ${nuscenes_result_dir} \
    --data_folder ${nuscenes2hz_data_dir}

4. How do I evaluate NuScenes
5. Look forward to your help

Hi, I download the validation_ground_truth_objects_gt.bin file from waymo but return error Traceback (most recent call last): File "gt_bin_decode.py", line 129, in main(args.file_path, out_folder, args.data_folder) File "gt_bin_decode.py", line 81, in main segment_name = segment_name_list[val_index] TypeError: list indices must be integers or slices, not NoneType When I try to decode the ground truth .bin file