{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/video-trajectory-classification-and-anomaly","title":"Video Trajectory Classification and Anomaly Detection Using Hybrid CNN-VAE","arxiv_id":"1812.07203","date":"2018-12-18","proceeding":null,"authors":["Santhosh Kelathodi Kumaran","Debi Prosad Dogra","Partha Pratim Roy","Adway Mitra"],"abstract":"Classifying time series data using neural networks is a challenging problem\nwhen the length of the data varies. Video object trajectories, which are key to\nmany of the visual surveillance applications, are often found to be of varying\nlength. If such trajectories are used to understand the behavior (normal or\nanomalous) of moving objects, they need to be represented correctly. In this\npaper, we propose video object trajectory classification and anomaly detection\nusing a hybrid Convolutional Neural Network (CNN) and Variational Autoencoder\n(VAE) architecture. First, we introduce a high level representation of object\ntrajectories using color gradient form. In the next stage, a semi-supervised\nway to annotate moving object trajectories extracted using Temporal Unknown\nIncremental Clustering (TUIC), has been applied for trajectory class labeling.\nAnomalous trajectories are separated using t-Distributed Stochastic Neighbor\nEmbedding (t-SNE). Finally, a hybrid CNN-VAE architecture has been used for\ntrajectory classification and anomaly detection. The results obtained using\npublicly available surveillance video datasets reveal that the proposed method\ncan successfully identify some of the important traffic anomalies such as\nvehicles not following lane driving, sudden speed variations, abrupt\ntermination of vehicle movement, and vehicles moving in wrong directions. The\nproposed method is able to detect above anomalies at higher accuracy as\ncompared to existing anomaly detection methods.","url_abs":"http://arxiv.org/abs/1812.07203v1","url_pdf":"http://arxiv.org/pdf/1812.07203v1.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"video-trajectory-classification-and-anomaly","repo_url":"https://github.com/lisaong/hss","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":null},{"paper_slug":"video-trajectory-classification-and-anomaly","repo_url":"https://github.com/santhoshkelathodi/CNN-VAE","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":null},{"paper_slug":"video-trajectory-classification-and-anomaly","repo_url":"https://github.com/santhoshkelathodi/CNN-VAE-based-Trajectory-Classification-and-Anomaly-Detection","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":null}],"tasks":[{"task_slug":"anomaly-detection","task_name":"Anomaly Detection"},{"task_slug":"classification-1","task_name":"Classification"},{"task_slug":"clustering","task_name":"Clustering"},{"task_slug":"classification","task_name":"General Classification"},{"task_slug":"object","task_name":"Object"},{"task_slug":"time-series-1","task_name":"Time Series"},{"task_slug":"time-series","task_name":"Time Series Analysis"}],"methods":[{"method_slug":"speed","method_name":"SPEED"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":null,"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}