search

SysCV / qd-3dt

Official implementation of Monocular Quasi-Dense 3D Object Tracking, TPAMI 2022
https://eborboihuc.github.io/QD-3DT/
BSD 3-Clause "New" or "Revised" License
517 stars 98 forks source link

eval_det_nusc.txt: No such file or directory #40

Closed LittleFlyFish closed 1 year ago

LittleFlyFish commented 1 year ago

I follow the instruction to install the environment and download the nuscenes dataset successfully. when I try to run this line:

image

I got the following errors:

it said lacking the file:

eval_det_nusc.txt eval_mot_nusc.txt eval_mot_02_nusc.txt

However, I cannot find it. Where could I find these files and how could I generate them?

Many thanks!

RoyYang0714 commented 1 year ago

Hi, can you show the full error msg? If you miss those files, then it might be an error in the model inference. Thus when it start to execute the nuscenes evaluation, it got the error.

LittleFlyFish commented 1 year ago

Hi Roy,

Thank you for your swift reply!

The error message is:

Best Regards,

Yanran

On Thu, Jun 1, 2023 at 6:54 PM Roy Yang @.***> wrote:

Hi, can you show the full error msg? If you miss those files, then it might be an error in the model inference. Thus when it start to execute the nuscenes evaluation, it got the error.

— Reply to this email directly, view it on GitHub https://github.com/SysCV/qd-3dt/issues/40#issuecomment-1571820594, or unsubscribe https://github.com/notifications/unsubscribe-auth/AGHDUUVZLMLYUMJWVZ5BFOTXJBYGRANCNFSM6AAAAAAYWWYABY . You are receiving this because you authored the thread.Message ID: @.***>

RoyYang0714 commented 1 year ago

It looks like you execute the scripts under $HOME, but our scripts should be executed under qd-3dt repo root.

LittleFlyFish commented 1 year ago

Hi Roy,

Thank you for the reply. I have checked that I have modified the file path so it shouldn't be a problem if I change the ./ dictionary. I think the main problem is:

work_dirs/Nusc/quasi_r101_dcn_3dmatch_multibranch_conv_dep_dim_cen_clsrot_sep_aug_confidence_scale_no_filter/output_val_box3d_deep_depth_motion_lstm_3dcen/eval_det_nusc.txt: No such file or directory

I double-checked the eval_det_nusc.txt file is not existing or generated. Could you help me check if your folder contains this file or which python script generate this file? Then I can dig into the problem better.

Thank you very much for your generous help and precious time! : )

On Fri, Jun 2, 2023 at 11:53 PM Roy Yang @.***> wrote:

It looks like you execute the scripts under $HOME, but our scripts should be executed under qd-3dt repo root.

— Reply to this email directly, view it on GitHub https://github.com/SysCV/qd-3dt/issues/40#issuecomment-1573957979, or unsubscribe https://github.com/notifications/unsubscribe-auth/AGHDUUWFYFBRG2WIVCJES4TXJID65ANCNFSM6AAAAAAYWWYABY . You are receiving this because you authored the thread.Message ID: @.***>

RoyYang0714 commented 1 year ago

I see. Then do you find output.json or output.pkl file under your work_dirs/Nusc/quasi_r101_dcn_3dmatch_multibranch_conv_dep_dim_cen_clsrot_sep_aug_confidence_scale_no_filter/output_val_box3d_deep_depth_motion_lstm_3dcen? If no, then you need to check whether your inference is completed or interrupted by bugs / env error.

LittleFlyFish commented 1 year ago

I see. Then do you find output.json or output.pkl file under your work_dirs/Nusc/quasi_r101_dcn_3dmatch_multibranch_conv_dep_dim_cen_clsrot_sep_aug_confidence_scale_no_filter/output_val_box3d_deep_depth_motion_lstm_3dcen? If no, then you need to check whether your inference is completed or interrupted by bugs / env error.

I have checked the file: Under work_dirs/Nusc/quasi_r101_dcn_3dmatch_multibranch_conv_dep_dim_cen_clsrot_sep_aug_confidence_scale_no_filter

there is no file of output_val_box3d_deep_depth_motion_lstm_3dcen or output.json and output.pkl

Only a file:

latest.pth

What could the problem be?

LittleFlyFish commented 1 year ago

I see. Then do you find output.json or output.pkl file under your work_dirs/Nusc/quasi_r101_dcn_3dmatch_multibranch_conv_dep_dim_cen_clsrot_sep_aug_confidence_scale_no_filter/output_val_box3d_deep_depth_motion_lstm_3dcen? If no, then you need to check whether your inference is completed or interrupted by bugs / env error.

Hi Roy,

I have finally solved the problem! Actually I should add one line in the test_eval_video_exp.py

args.add_ablation_exp = ['all']

at line 958

the variable add_ablation_exp should has 'all' before running the code to generate the files.

Hope this helps!