Detecting and counting on road vehicles is a key task in intelligent transport management and surveillance systems. The applicability lies both in urban and highway traffic monitoring and control, particularly in difficult weather and traffic conditions. In the past, the task has been performed through data acquired from sensors and conventional image processing toolbox. However, with the advent of emerging deep learning based smart computer vision systems the task has become computationally efficient and reliable. The data acquired from road mounted surveillance cameras can be used to train models which can detect and track on road vehicles for smart traffic analysis and handling problems such as traffic congestion particularly in harsh weather conditions where there are poor visibility issues because of low illumination and blurring. Different vehicle detection algorithms focusing the same issue deal only with on or two specific conditions. In this research, we address detecting vehicles in a scene in multiple weather scenarios including haze, dust and sandstorms, snowy and rainy weather both in day and nighttime. The proposed architecture uses CSPDarknet53 as baseline architecture modified with spatial pyramid pooling (SPP-NET) layer and reduced Batch Normalization layers. We also augment the DAWN Dataset with different techniques including Hue, Saturation, Exposure, Brightness, Darkness, Blur and Noise. This not only increases the size of the dataset but also make the detection more challenging. The model obtained mean average precision of 81% during training and detected smallest vehicle present in the image

Retinoblastoma is a rare and aggressive form of childhood eye cancer that requires prompt diagnosis and treatment to prevent vision loss and even death. Deep learning models have shown promising results in detecting retinoblastoma from fundus images, but their decision-making process is often considered a “black box” that lacks transparency and interpretability. In this project, we explore the use of LIME and SHAP, two popular explainable AI techniques, to generate local and global explanations for a deep learning model based on InceptionV3 architecture trained on retinoblastoma and non-retinoblastoma fundus images. We collected and labeled a dataset of 400 retinoblastoma and 400 non-retinoblastoma images, split it into training, validation, and test sets, and trained the model using transfer learning from the pre-trained InceptionV3 model. We then applied LIME and SHAP to generate explanations for the model’s predictions on the validation and test sets. Our results demonstrate that LIME and SHAP can effectively identify the regions and features in the input images that contribute the most to the model’s predictions, providing valuable insights into the decision-making process of the deep learning model. In addition, the use of InceptionV3 architecture with spatial attention mechanism achieved high accuracy of 97% on the test set, indicating the potential of combining deep learning and explainable AI for improving retinoblastoma diagnosis and treatment.

No abstract

Pressure ulcers are significant healthcare concerns affecting millions of people worldwide, particularly those with limited mobility. Early detection and classification of pressure ulcers are crucial in preventing their progression and reducing associated morbidity and mortality. In this work, we present a novel approach that uses YOLOv5, an advanced and robust object detection model, to detect and classify pressure ulcers into four stages and non-pressure ulcers. We also utilize data augmentation techniques to expand our dataset and strengthen the resilience of our model. Our approach shows promising results, achieving an overall mean average precision of 76.9% and class-specific mAP50 values ranging from 66% to 99.5%. Compared to previous studies that primarily utilize CNN-based algorithms, our approach provides a more efficient and accurate solution for the detection and classification of pressure ulcers. The successful implementation of our approach has the potential to improve the early detection and treatment of pressure ulcers, resulting in better patient outcomes and reduced healthcare costs.

In this research, we address the problem of accurately predicting lane-change maneuvers on highways. Lane-change maneuvers are a critical aspect of highway safety and traffic flow, and the accurate prediction of these maneuvers can have significant implications for both. However, current methods for lane-change prediction are limited in their ability to handle naturalistic driving scenarios and often require large amounts of labeled data. Our proposed model uses a bidirectional long short-term memory (BiLSTM) network to analyze naturalistic vehicle trajectories recorded from multiple sensors on German highways. To handle the temporal aspect of vehicle behavior, we utilized a sliding window approach, considering both the preceding and following vehicles’ trajectories. To tackle class imbalances in the data, we introduced rolling mean computed weights. Our extensive feature engineering process resulted in a comprehensive feature set to train the model. The proposed model fills the gap in the state-of-the-art lane change prediction methods and can be applied in advanced driver assistance systems (ADAS) and autonomous driving systems. Our results show that the BiLSTM-based approach with the sliding window technique effectively predicts lane changes with 86% test accuracy and a test loss of 0.325 by considering the context of the input data in both the past and future. The F1 score of 0.52, precision of 0.41, recall of 0.75, accuracy of 0.86, and AUC of 0.81 also demonstrate the model’s high ability to distinguish between the two target classes. Furthermore, the model achieved an accuracy of 83.65% with a loss value of 0.3306 on the other half of the data samples, and the validation accuracy was observed to improve over these epochs, reaching the highest validation accuracy of 92.53%. The F1 score of 0.51, precision of 0.36, recall of 0.89, accuracy of 0.82, and AUC of 0.85 on this data sample also demonstrate the model’s strong ability to identify both positive and negative classes. Overall, our proposed approach outperforms existing methods and can significantly contribute to improving highway safety and traffic flow.