Goal: Research the Impact of Imbalanced and Balanced Maritime Code Flag Datasets on the Performance of One-Stage Image Detectors (YOLO family, SSD).
Main metrics:
- intersection over union (IoU),
- precision and recall,
- average precision (AP),
- mean average precision (mAP),
- F1 score (trade-off between precision and recall).
YOLO detectors:
- YOLOv11 (Yolov11m ~ Medium version) - latest stable version (REQUIRED FOR RESEARCH),
- YOLOv8 (Yolov8m ~ Medium version) (REQUIRED FOR RESEARCH).
SSD detectors:
- SSD300 - latest stable version (REQUIRED FOR RESEARCH),
- SSD512 - bigger resoultions (ADDITIONAL FOR RESEARCH).
├── documentation <- UML diagrams and configuration
├── balancers <- Package with balancers and utils
│ ├── __init__.py <- Package identicator
│ ├── smote.py <- SMOTE balancer (interpolation)
│ ├── adasyn.py <- ADASYN balancer (interpolation)
│ ├── augmentation.py <- Augmentation balancer (augmenting images like rotations, etc.)
│ ├── autoencoder.py <- Autoencoder balancer (learning needed!)
│ ├── dgan.py <- DGAN balancer (learning needed!)
│ ├── balancer.py <- General balancer with all balancers (aggregating all of the above)
│ ├── annotations.py <- Annotations module
│ └── configuration_reader.py <- Balancer configuration reader
├── maritime-flags-dataset <- Source and balanced flags (A-Z)
│ ├── ADASYN_balanced_flags <- Balanced flags by using ADASYN balancer
│ ├── SMOTE_balanced_flags <- Balanced flags by using SMOTE balancer
│ ├── AUGMENTATION_balanced_flags <- Balanced flags by using Augmentation balancer
│ ├── DGAN_balanced_flags <- Balanced flags by using DGAN balancer
│ ├── AE_balanced_flags <- Balanced flags by using Autoencoder balancer
│ ├── combined_flags <- Combined/test images
│ ├── two_flags <- Balanced two flags (A and B) per 1000 images
│ └── imbalanced_flags <- Source folder with imbalanced flags
├── datasets <- YOLO formatted datasets (detector by default sees this category!)
│ ├── yolo-maritime-flags-dataset (A-Z)
│ ├── images
│ ├── train <- Training images (.jpg)
│ ├── val <- Validation images (.jpg)
│ └── test <- Testing images (.jpg)
│ └── labels
│ ├── train <- Training labels (.txt)
│ ├── val <- Validation labels (.txt)
│ └── test <- Testing labels (.txt)
│ └── cross-validation-yolo-formatted-maritime-flags (A-Z)
│ ├── images
│ ├── fold_1 <- First fold with images (.jpg)
| ├── train <- Training images (.jpg)
| └── val <- Validation images (.jpg)
│ ├── ... <- ... fold with images (.jpg)
| ├── train <- Training images (.jpg)
| └── val <- Validation images (.jpg)
│ └── fold_n <- N-fold with images (.jpg)
| ├── train <- Training images (.jpg)
| └── val <- Validation images (.jpg)
│ └── labels
│ ├── fold_1 <- First fold with labels (.txt)
| ├── train <- Training labels (.txt)
| └── val <- Validation labels (.txt)
│ ├── ... <- ... fold with labels (.txt)
| ├── train <- Training labels (.txt)
| └── val <- Validation labels (.txt)
│ └── fold_n <- N-fold with labels (.txt)
| ├── train <- Training labels (.txt)
| └── val <- Validation labels (.txt)
├── .gitignore <- Ignores venv_environment directory to be pushed (VENV)
├── test_packages.py <- Testing loading all necessaries packages like Torch (VENV)
├── python_3.11_venv_requirements.txt <- List for venv with all used packages (VENV)
├── balance.py <- Balancing dataset by using balancers package (BALANCING)
├── balancer_configuration.json <- Balancer configuration
├── detection.py <- Training and testing yolo detector with balanced/imbalanced data (EVALUATING)
├── yolo_detector.py <- YOLO detector (DETECTING)
├── yolo_data.yaml <- YOLO data configuration (traing and testing)
├── fold_1_dataset.yaml <- YOLO first k-fold data configuration (K-cross validation)
├── ... <- YOLO ... k-fold data configuration (K-cross validation)
└── fold_n_dataset.yaml <- YOLO n k-fold data configuration (K-cross validation)
