This repository contains code and models for a drowsiness detection system. Below is a brief description of the files and folders:
main_multi.py: Main drowsiness code.eye.py: Helper class used for eye-based feature calculation, used in tandem withmain_multi.py.final.pkl: XGBoost model for drowsiness detection.
-
deep_learning/: Folder containing deep learning models and associated training scripts. -
feature_selection/: Folder containing various files for feature selection:feature_visualization.py: Visualization of features graph with their output.feature_combination.py: Combination of features tested for various models.feature_selection.py: Various feature selection methods.
-
hyperparameter_tuning/: Folder containing hyperparameter tuning scripts:random_search_cv.py: Hyperparameter tuning using RandomSearchCV.hyperopt.py: Hyperparameter tuning using Hyperopt.stepwise_xgboost.py: Stepwise approach using XGBoost for hyperparameter tuning (GROUP 1: max_depth, min_child_weight, GROUP 2: subsample, colsample_bytree, GROUP 3: learning_rate, num_boost_round).
The dataset used was the UTARLDD found here: https://www.kaggle.com/datasets/rishab260/uta-reallife-drowsiness-dataset
To run the code, execute main_multi.py. The first 200 frames (corresponds to ~ 5 - 10 seconds depending on the system) are used for calibrating. During this time, stare into the camera normally and blink a maximum of one time.
The following table displays the accuracy of various models:
| Model | Accuracy |
|---|---|
| XGBoost | 90% |
| Bagging Classifier | 89% |
| Decision Tree | 87% |
| LGM Classifier | 87% |
| Gradient Boost | 84% |
| Random Forest | 81% |
| K Neighbours | 80% |
| Extra Tree | 78% |
| AdaBoost | 72% |
| Naïve Bayes | 41% |
| SVM | 41% |
| Logistic Regression | 33% |
| LSTM | 72% |
| LSTM - Stacked | 75% |
| RNN | 69% |
| RNN - Stacked | 72% |