YOLOv1-Circle Training Script

This project implements a YOLOv1 model using circular bounding boxes for object detection. The model is trained on custom datasets and leverages PyTorch for deep learning.

Requirements

• Python 3.8+
• PyTorch
• torchvision
• CUDA (optional, for GPU acceleration)
• Colab environment (for the given paths)

Installation

1. Install dependencies by running the following command:

   pip install torch torchvision

2. Clone the repository and place your datasets in the specified directories or update the paths accordingly.

Training Instructions

1. Update Paths:

   • Modify the following paths in the script to point to your own directories:
     • PATH_TO_SAVE_MODEL: Path to save the trained YOLO model.
     • PATH_TO_SAVE_CHECKPOINT: Path to save model checkpoints.
     • DST_PATH_END_CSV: Path to save the CSV file containing training loss and AP (Average Precision).
     • ARH_PATH_0, ARH_PATH_1, ARH_PATH_2: Paths to your dataset archives. These archives will be unpacked during training.

2. Training Configuration:

   • The model hyperparameters (such as batch size, learning rate, and epochs) can be adjusted in the script. The default settings are:
     • BATCH_SIZE = 64
     • LEARNING_RATE = 1e-5
     • EPOCHS = 250
   • Other important settings like image size, number of classes, and stride are also configurable in the script.

3. Run the Script:

   • Ensure that the necessary datasets are available, and the paths to the datasets are correctly specified.

   • To start the training, execute the following command:

     python train_yolov1_circle.py

4. Model Checkpoints:

   • The model will save checkpoints every 5 epochs (or as specified). If LOAD_MODEL = True, the script will attempt to load the last checkpoint to resume training.

5. CSV Logging:

   • During training, the loss and AP for each epoch will be logged into the CSV file specified by DST_PATH_END_CSV.

Example Paths

PATH_TO_SAVE_MODEL = '/content/drive/MyDrive/Models/yoloV1_circle.pth'
PATH_TO_SAVE_CHECKPOINT = '/content/drive/MyDrive/Models/checkpoint_yoloV1_circle.pth'
DST_PATH_END_CSV = '/content/drive/MyDrive/Results/yoloV1_loss_ap.csv'
ARH_PATH_0 = '/content/drive/MyDrive/Datasets/dataset1.zip'
ARH_PATH_1 = '/content/drive/MyDrive/Datasets/dataset2.zip'
ARH_PATH_2 = '/content/drive/MyDrive/Datasets/dataset3.zip'

Make sure to modify the paths above to match the location of your files and directories.