Compound effects of top-down and bottom-up influences on visual attention during action recognition

摘要

The limited visual and computational resources available during the perception of a human action makes a visual attention mechanism essential.In this paper we propose an attention mechanism that combines the saliency of top-down (or goaldirected) elements, based on multiple hypotheses about the demonstrated action, with the saliency of bottom-up (or stimulus-driven) components.Furthermore, we use the bottom-up part to initialise the top-down, hence resulting in a selection of the behaviours that rightly require the limited computational resources.This attention mechanism is then combined with an action understanding model and implemented on a robot, where we examine its performance during the observation of object-directed human actions.