We used semi-supervised techniques to detect traffic incident trajectories from the cameras. Our proposed approach for trajectory classification is based on semi-supervised parameter estimation using maximum-likelihood (ML) estimation. Results show that approximately 14% improvement in trajectory classification can be achieved using the proposed approach compared to baseline algorithms. 
This paper used Twitter data to extract sentiments of people during the Phase 1 and Phase 2 of the odd-even policy implemented by the Delhi government to curb the air pollution and improve traffic flow. Public sentiments were found to mostly turn negative during the later stage of the Phase 2 which indicates fading away of the public enthusiasm and positiveness towards the policy during the later stages of the policy implementation. 
In this study we used two different deep learning techniques, you only look once (YOLO) and deep convolution neural network (DCNN), and a shallow algorithm, support vector machine (SVM), to detect traffic congestion from camera images. YOLO and DCCN achieved 91.5% and 90.2% accuracy, respectively, whereas SVM’s accuracy was 85.2%. 