Abstract: With the increasing number and variety of camera installations, unsupervised methods that learn typical activities have become popular for anomaly detection. In this thesis, we con-sider recent methods based on temporal probabilistic models and improve them in multiple ways. Our contributions are the following: (i) we integrate the low level processing and the temporal activity modeling, showing how this feedback improves the overall quality of the captured information, (ii) we show how the same approach can be taken to do hierarchical multi-camera processing, (iii) we use spatial analysis of the anomalies both to perform local anomaly detection and to frame automatically the detected anomalies. We illustrate the approach on both tra?c data and videos coming from a metro station. We also investigate the application of topic models in Brain Computing Interfaces for Mental Task classi?cation. We observe a classi?cation accuracy of up to 68% for four Mental Tasks on individual subjects.
Adviser: Sanjay Rajopadhye Co-Adviser: Chuck Anderson Non-ECE Member: Anton Bohm Member 3: Sudeep Pasricha Addional Members:
Publications:  Thiyagarajan Chockalingam, Remi Emonet, Jean-Marc Odobez. "Localized Anomaly Detection via Hierarchical Integrated Activity Discovery." 10-th edition of the IEEE International Conference on Advanced Video and Signal-Based Surveillance (AVSS), 2013
Program of Study: CS453 CS575 CS787 ECE560 ECE561 ECE699 GRAD510 MATH560