thigiacmaytinh / beverage_yolov7

Install package and library

• Download Dataset from Robotflow: Đã download trong folder drinks-1
• Install Python 3.7.3 AMD64
• Install Enviroment:

pip install -r requirements.txt

• Install CUDA 11.3 and CuDNN 8.2.1.32

• Install Pytorch

pip install torch==1.12.1+cu113 torchvision==0.13.1+cu113 torchaudio==0.12.1 --extra-index-url https://download.pytorch.org/whl/cu113

• Install Visual Studio Code

• Download file pretrain

https://github.com/WongKinYiu/yolov7/releases/download/v0.1/yolov7.pt

Training:

Training bằng file notebook yolov7_training.ipynb

Ghi chú

Chạy file client.py

• Leak mem -> comment dòng x[:, 5:] *= x[:, 4:5] # conf = obj_conf * cls_conf trong file ultis/general.py (dòng 648->652)

Điều chỉnh tham số truyền vào class coke trong file client.py

param weights: the path to the weights file param source: the path to the video file param conf: the confidence threshold for the bounding boxes param imgsz: The size of the image to be processed, defaults to 640 (optional) param device: the GPU device to use, defaults to 0 (optional) param iou_thres: The IoU threshold for non-maximum suppression

Ảnh sau khi Detect sẽ được lưu trong run/detect/exp