Problems of running gmtracker_app.py

[NanH5837]

Hello,when I tried to run python gmtracker_app.py --sequence_dir data/MOT17/test/MOT17-01 --detection_file npy/npytest_tracktor/MOT17-01-DPM.npy --checkpoint_dir experiments/static/params/0001 --max_age 100 --reid_thr 0.6 --output_file results/test/MOT17-01-DPM.txt in Pycharm,it show errors as follows:

`torch.Size([2171, 512]) Loading model parameters from experiments/static/params/0001/params.pt torch.Size([2171, 512]) Processing MOT17-01 frame 00001 Nothing to match at this level Processing MOT17-01 frame 00002 Nothing to match at this level Processing MOT17-01 frame 00003

Process finished with exit code 139 (interrupted by signal 11: SIGSEGV)`

I have no idea what causes the issue.Can you help me?

[jiaweihe1996]

How much memory do you have?

[NanH5837]

How much memory do you have?

32G

[jiaweihe1996]

Could you use pdb to locate the line where reporting the error? And do you have GPUs? If not, please use CPU version of the SCS solver, i.e., use prob.solve(solver=cp.SCS) instead of https://github.com/jiaweihe1996/GMTracker/blob/7f0e9c1d46871b609b0241af71b9e8d619e1bc0f/GMMOT/gmatching.py#L91 and https://github.com/jiaweihe1996/GMTracker/blob/7f0e9c1d46871b609b0241af71b9e8d619e1bc0f/GMMOT/gmatching.py#L107

[NanH5837]

Could you use pdb to locate the line where reporting the error? And do you have GPUs? If not, please use CPU version of the SCS solver, i.e., use prob.solve(solver=cp.SCS) instead of

https://github.com/jiaweihe1996/GMTracker/blob/7f0e9c1d46871b609b0241af71b9e8d619e1bc0f/GMMOT/gmatching.py#L91

and https://github.com/jiaweihe1996/GMTracker/blob/7f0e9c1d46871b609b0241af71b9e8d619e1bc0f/GMMOT/gmatching.py#L107 Thank you.My GPU is GTX 1080 .I haven't tried pdb yet and I'll try it and get back to you

[NanH5837]

Could you use pdb to locate the line where reporting the error? And do you have GPUs? If not, please use CPU version of the SCS solver, i.e., use prob.solve(solver=cp.SCS) instead of

https://github.com/jiaweihe1996/GMTracker/blob/7f0e9c1d46871b609b0241af71b9e8d619e1bc0f/GMMOT/gmatching.py#L91

and https://github.com/jiaweihe1996/GMTracker/blob/7f0e9c1d46871b609b0241af71b9e8d619e1bc0f/GMMOT/gmatching.py#L107

I tried to set pdb as follows: pdb.set_trace() if Mp.shape[0] > Mp.shape[1]: pdb.set_trace()

the issue might be in gmatching.py:78 and it shows:

`torch.Size([2171, 512]) Loading model parameters from experiments/static/params/0001/params.pt torch.Size([2171, 512]) Processing MOT17-01 frame 00001 Nothing to match at this level Processing MOT17-01 frame 00002 Nothing to match at this level Processing MOT17-01 frame 00003

/home/xxx/MOT/GMTracker/GMMOT/gmatching.py(81)forward() -> if Mp.shape[0] > Mp.shape[1]: (Pdb) c

Process finished with exit code 139 (interrupted by signal 11: SIGSEGV)`

I can't solve it

[jiaweihe1996]

In MOT17-01-DPM frame 00003, Mp.shape[0] > Mp.shape[1] is false, and the program goes to https://github.com/jiaweihe1996/GMTracker/blob/7f0e9c1d46871b609b0241af71b9e8d619e1bc0f/GMMOT/gmatching.py#L95 So, you should set trace under the line 95 and run line by line to locate the error.

[NanH5837]

In MOT17-01-DPM frame 00003, Mp.shape[0] > Mp.shape[1] is false, and the program goes to

https://github.com/jiaweihe1996/GMTracker/blob/7f0e9c1d46871b609b0241af71b9e8d619e1bc0f/GMMOT/gmatching.py#L95

So, you should set trace under the line 95 and run line by line to locate the error.

The error should be in gmatching.py:107 prob.solve(solver=cp.SCS, gpu=True, use_indirect=True) How can I solve it?

[jiaweihe1996]

Could you pass the SCS solver test?

cd scs-python
python test/solve_random_cone_prob.py

If not, please reinstall SCS. Or pip install scs, and use cpu version of SCS solver.

[NanH5837]

Could you pass the SCS solver test?

cd scs-python
python test/solve_random_cone_prob.py

If not, please reinstall SCS. Or pip install scs, and use cpu version of SCS solver.

Thanks very much .I use cpu version of SCS solver to solve this error.