[AICUP 2024] Competition-2024-跨相機多目標車輛追蹤競賽

https://tbrain.trendmicro.com.tw/Competitions/Details/33
Extremely low frame-rate (1 fps) video object tracking challenge

2nd place in 跨相機多目標車輛追蹤競賽模型組 and 執行效能獎

https://github.com/user-attachments/assets/20fc4543-e6ec-4d56-9a64-8a7bcaa828c3

https://github.com/user-attachments/assets/a8ea9150-8dc6-4d99-8f0b-ba2ed34229f0

Setup with Conda

Step 1. Create Conda environment and install pytorch.

conda create -n botsort python=3.7
conda activate botsort

Step 2. Install torch and matched torchvision from pytorch.org.
The code was tested using torch 1.10.1+cu102 and torchvision==0.11.2

Step 3. Fork this Repository and clone your Repository to your device

Step 4. Install numpy first!!

pip install numpy

Step 5. Install requirement.txt

pip install -r requirement.txt

Step 6. Install pycocotools.

pip install cython; pip3 install 'git+https://github.com/cocodataset/cocoapi.git#subdirectory=PythonAPI'

Step 7. Others

# Cython-bbox
pip install cython_bbox

# faiss cpu / gpu
pip install faiss-cpu
pip install faiss-gpu

ReID Model

For training the ReID, detection patches must be generated as follows:

python fast_reid/datasets/generate_AICUP_patches.py --data_path <dataets_dir>/AI_CUP_MCMOT_dataset/train

Train ReID MODEL

python fast_reid/tools/train_net.py --config-file fast_reid/configs/AICUP/bagtricks_R50-ibn.yml MODEL.DEVICE "cuda:0"

The training results are stored by default in logs/AICUP/bagtricks_R50-ibn. The storage location and model hyperparameters can be modified in fast_reid/configs/AICUP/bagtricks_R50-ibn.yml. You can refer to fast_reid/fastreid/config/defaults.py to find out which hyperparameters can be modified.

YOLOv7 Model

run the yolov7/tools/AICUP_to_YOLOv7.py by the following command:

cd <BoT-SORT_dir>
python yolov7/tools/AICUP_to_YOLOv7.py --AICUP_dir datasets/AI_CUP_MCMOT_dataset/train --YOLOv7_dir datasets/AI_CUP_MCMOT_dataset/yolo

The file tree after conversion by AICUP_to_YOLOv7.py is as follows:

/datasets/AI_CUP_MCMOT_dataset/yolo
    ├── train
    │   ├── images
    │   │   ├── 0902_150000_151900_0_00001.jpg (Date_StartTime_EndTime_CamID_FrameNum)
    │   │   ├── 0902_150000_151900_0_00002.jpg
    │   │   ├── ...
    │   │   ├── 0902_150000_151900_7_00001.jpg
    │   │   ├── 0902_150000_151900_7_00002.jpg
    │   │   ├── ...
    │   └── labels
    │   │   ├── 0902_150000_151900_0_00001.txt (Date_StartTime_EndTime_CamID_FrameNum)
    │   │   ├── 0902_150000_151900_0_00002.txt
    │   │   ├── ...
    │   │   ├── 0902_150000_151900_7_00001.txt
    │   │   ├── 0902_150000_151900_7_00002.txt
    │   │   ├── ...
    ├── valid
    │   ├── images
    │   │   ├── 1015_190000_191900_0_00001.jpg (Date_StartTime_EndTime_CamID_FrameNum)
    │   │   ├── 1015_190000_191900_0_00002.jpg
    │   │   ├── ...
    │   │   ├── 1015_190000_191900_7_00001.jpg
    │   │   ├── 1015_190000_191900_7_00002.jpg
    │   │   ├── ...
    │   └── labels
    │   │   ├── 1015_190000_191900_0_00001.txt (Date_StartTime_EndTime_CamID_FrameNum)
    │   │   ├── 1015_190000_191900_0_00002.txt
    │   │   ├── ...
    │   │   ├── 1015_190000_191900_7_00001.txt
    │   │   ├── 1015_190000_191900_7_00002.txt
    │   │   ├── ...

the trained models in 'pretrained' folder as follows:

<BoT-SORT_dir>/pretrained

Final yolov7 trained weight : pretrained/yolov7-w6-AICUP7_049.pt. Yolov7 pretrained model : pretrained/yolov7-w6_training.pt.

Fine-tune YOLOv7 for AICUP

The dataset path is configured in yolov7/data/AICUP.yaml.
The model architecture can be configured in yolov7/cfg/training/yolov7-AICUP.yaml.
Training hyperparameters are configured in yolov7/data/hyp.scratch.custom.yaml (default is yolov7/data/hyp.scratch.p5.yaml).
pretrained model : pretrained/yolov7-w6_training.pt from (https://github.com/WongKinYiu/yolov7/releases/download/v0.1/)

python yolov7/train_aux.py --device "0" --batch-size 4 --epochs 50 --data yolov7/data/AICUP.yaml --img 1280 1280 --cfg yolov7/cfg/training/yolov7-w6-AICUP.yaml --weights 'pretrained/yolov7-w6_training.pt' --name yolov7-w6-AICUP --hyp data/hyp.scratch.custom.yaml

The training results will be saved by default at runs/train.

Tracking and creating the submission file for AICUP

If you want to track all <timestamps>testdata the same as submit, you can execute the bash file we provided and clean cnt.txt file to 0. Make sure the ID start from 0

sh tools/track_all_timestamps2.sh --weights pretrained/yolov7-w6-AICUP7_049.pt --source-dir ./<test_data_dir>/32_33_AI_CUP_testdataset/AI_CUP_testdata/images --device "0" --fast-reid-config "fast_reid/configs/AICUP/bagtricks_R50-ibn.yml" --fast-reid-weights "logs/AICUP_BOT_resnet/bagtricks_R50-ibn/model_0058.pth"

The submission file and visualized images will be saved by default at runs/submit/<timestamp>.

Track the spesific timestamp

If you want to track specific <timestamps> video data, you can execute the bash as following

python tools/mc_demo_yolov7_day_night_submit.py --weights "$WEIGHTS" --source "$folder" --device "$DEVICE" --name "$timestamp" --fuse-score --agnostic-nms --with-reid --fast-reid-config "$FAST_REID_CONFIG" --fast-reid-weights "$FAST_REID_WEIGHTS"

pdway53 / AICUP_ReID_Project

readme