Face Detection and Tracking for Activity Recognition

A. Kuzmanic Skelin, M. Bonkovic, and V. Papic (Croatia)


ace detection, face tracking, activity recognition


In this paper we present an approach to human activity recognition in an indoor environments. It combines Adaboost face detection with standard tracking methods – continuously adaptive mean-shift (CAMShift) and Kalman filtering. The output results are used to describe simple events of human activities in video using Finite State Machine. The system has been tested on several video sequences and we provide a summary of results.

Important Links:

Go Back