Algorithm Used in Video Event Recognition & Classification with Hierarchical Modeling

Video media data is easier to transmit, produce, and distribute over the world. There has been a lot of information lost since advancements in video analysis and processing have lagged behind the data's exponential increase. The necessity for effective recovery of video data content is, thus, pressing. Because there are so many potential classifications, identifying and labeling occurrences may be difficult. In order to help human operators keep an eye on everything, researchers have built algorithms to better understand the semantic content of videos so they can better spot potential threats. The Yolo method for object detection and the hierarchical context model for event recognition are both explored in this work using a common surveillance video dataset. We hope that by employing these methods, we can enhance the efficiency with which video data material is retrieved by better identifying and classifying occurrences.