Deep learning (DL) architectures are becoming increasingly popular in modern traffic systems and self-driven vehicles owing to their high efficiency and accuracy. Emerging technological advancements and the availability of large databases have made a favorable impact on such improvements. In this study, we present a traffic sign recognition system based on novel DL architectures, trained and tested on a locally collected traffic sign database. Our approach includes two stages; traffic sign identification from live video feed, and classification of each sign. The sign identification model was implemented with YOLO architecture and the classification model was implemented with Xception architecture. The input video feed for these models were collected using dashboard camera recordings. The classification model has been trained with the German Traffic Sign Recognition Benchmark dataset as well for comparison. Final accuracy of classification for the local dataset was 96.05% while the standard dataset has given an accuracy of 92.11%. The final model is a combination of the detection and classification algorithms and it is able to successfully detect and classify traffic signs from an input video feed within an average detection time of 4.5fps

Internet of Things (IoT) consists of group of devices which communicates information over private networks. One of the key challenges faced by IoT networks is the security breaches.  With the objective of automating the detection of possible security breaches in five categories, IoT traffic created with Message Queue Telemetry Transport (MQTT) protocol is analyzed. The five categories of cyber-attacks considered are brute force, denial of service (DoS), flooding, malformed data, and SlowITe attacks along with legitimate traffic.  The popular five machine learning (ML) models, LightGBM, Random Forest, MLP, AdaBoost, and Decision Tree Classifiers are trained to predict cyber-attacks. In traditional traffic analysis all the available features of MQTT traffic were utilized for the ML modeling and in this work, we challenge the practice by showing that automated feature selection improves the performance of the overall ML models. The average accuracy, precision, recall and the F1 score are used as performance evaluation metrics. It is observed that all models in average are able to achieve 90% of accuracy in classification, while MLP model is trained 10 times faster than the other models.  Further the optimal number of features for correct classification is identified as 10 features through Monte Carlo analysis. With the reduced features, it is possible to detect DoS, flooding, and  SlowITe attacks with more than 90% accuracy and precision. Yet, it is difficult to tell apart brute force and malformed data attacks.