After detecting the object in the camera image, by the transformation between the robot coordinate and the camera coordinate, the humanoid robot is able to operate in the error level of a 3 mm boundary. For example, the process of seeking and pushing an oven button with a radius of only 15 mm, is more exact due to the perfect recognition and the delicate calibration process. Here, it should be considered how the robot can execute a task when a certain object is not in its designated place but has moved to another location. The robot should decide whether to execute the task again, skip the task, or quit the task. The object’s location and orientation can be obtained by performing object pose tracking based on the model. This is done using a particle filter. This algorithm is used in many areas due to its suitable performance [16]-[19]. A conventional particle filter works as follows: