Learning Action in Time Sequential Images using Data Mining

1 Motivation Computer vision research has always been an active area of artificial intelligence. One of the basic goals of AI is to build agents that act intelligently or human-like. Reasoning, learning and language processing have contributed in reaching the goal of Turing test. Yet they provide only a partial solution as they lack the basic perceptual capabilities required to perceive and interact with the world rather than be fed by information externally. Agents should be capable of finding other static and non-static objects and their spatial relationship etc. Computer vision plays an important role here. Recently, Human motion analysis is receiving attention from computer vision researchers. This is due to the wide spectrum of applications that can benefit from successful human motion analysis such as surveillance, man-machine interfaces, content-based image storage and retrieval etc. For example, segmenting the parts of the human body in an image, tracking the movement of joints over an image sequence, and recovering the underlying 3D body structure is particularly useful for analysis of athletic performance as well as medical diagnostics. The capability to automatically monitor human activities using computers in security-sensitive areas such as airports, borders, and building lobbies is of great interest to the police and military. For example, consider the problem of finding whether a car avoids a check point or not. In case of human surveillance, the person may not be attentive over long periods and can miss some event. But automated surveillance is attentive for long durations without missing any such events. Computer vision can also be deployed in the area of detection of unusual events. With the development of digital libraries, the ability to automatically interpret video sequences will save tremendous human effort in sorting and retrieving images or video sequences using content-based queries. Other applications include building man-machine user interfaces, video conferencing etc.