MaochengHu 576cda45b8 first commit | 2 роки тому | |
---|---|---|
.. | ||
mot_online | 2 роки тому | |
README.md | 2 роки тому | |
byte_tracker.py | 2 роки тому | |
eval.py | 2 роки тому | |
evaluation.py | 2 роки тому | |
joint.py | 2 роки тому | |
motr.py | 2 роки тому | |
motr_det.py | 2 роки тому | |
transforms.py | 2 роки тому |
Step1.
git clone https://github.com/megvii-model/MOTR.git and install
replace https://github.com/megvii-model/MOTR/blob/main/datasets/joint.py
replace https://github.com/megvii-model/MOTR/blob/main/datasets/transforms.py
train
python3 -m torch.distributed.launch --nproc_per_node=8 \
--use_env main.py \
--meta_arch motr \
--dataset_file e2e_joint \
--epoch 50 \
--with_box_refine \
--lr_drop 40 \
--lr 2e-4 \
--lr_backbone 2e-5 \
--pretrained coco_model_final.pth \
--output_dir exps/e2e_motr_r50_mot17trainhalf \
--batch_size 1 \
--sample_mode 'random_interval' \
--sample_interval 10 \
--sampler_steps 10 20 30 \
--sampler_lengths 2 3 4 5 \
--update_query_pos \
--merger_dropout 0 \
--dropout 0 \
--random_drop 0.1 \
--fp_ratio 0.3 \
--query_interaction_layer 'QIM' \
--extra_track_attn \
--mot_path .
--data_txt_path_train ./datasets/data_path/mot17.half \
--data_txt_path_val ./datasets/data_path/mot17.val \
mot17.half and mot17.val are from https://github.com/ifzhang/FairMOT/tree/master/src/data
You can also download the MOTR model trained by us: google, baidu(code:t87h)
Step2.
replace https://github.com/megvii-model/MOTR/blob/main/util/evaluation.py
replace https://github.com/megvii-model/MOTR/blob/main/eval.py
replace https://github.com/megvii-model/MOTR/blob/main/models/motr.py
add byte_tracker.py to https://github.com/megvii-model/MOTR
add mot_online to https://github.com/megvii-model/MOTR
Step3.
val
python3 eval.py \
--meta_arch motr \
--dataset_file e2e_joint \
--epoch 200 \
--with_box_refine \
--lr_drop 100 \
--lr 2e-4 \
--lr_backbone 2e-5 \
--pretrained exps/e2e_motr_r50_mot17val/motr_final.pth \
--output_dir exps/e2e_motr_r50_mot17val \
--batch_size 1 \
--sample_mode 'random_interval' \
--sample_interval 10 \
--sampler_steps 50 90 120 \
--sampler_lengths 2 3 4 5 \
--update_query_pos \
--merger_dropout 0 \
--dropout 0 \
--random_drop 0.1 \
--fp_ratio 0.3 \
--query_interaction_layer 'QIM' \
--extra_track_attn \
--mot_path ./MOT17/images/train
--data_txt_path_train ./datasets/data_path/mot17.half \
--data_txt_path_val ./datasets/data_path/mot17.val \
--resume model_final.pth \
in Step2, replace https://github.com/megvii-model/MOTR/blob/main/models/motr.py by motr_det.py
others are the same as MOTR