PROJECT TITLE :

Event Oriented Dictionary Learning for Complex Event Detection

ABSTRACT:

Complicated event detection is a retrieval task with the goal of finding videos of a particular event in an exceedingly large-scale unconstrained Internet video archive, given example videos and text descriptions. Nowadays, different multimodal fusion schemes of low-level and high-level features are extensively investigated and evaluated for the complicated event detection task. However, how to effectively select the high-level semantic meaningful ideas from a giant pool to assist complex event detection is rarely studied within the literature. During this paper, we have a tendency to propose a novel strategy to automatically choose semantic meaningful ideas for the event detection task based on each the events-kit text descriptions and therefore the ideas high-level feature descriptions. Moreover, we tend to introduce a novel event oriented dictionary illustration primarily based on the selected semantic ideas. Toward this goal, we have a tendency to leverage training pictures (frames) of selected ideas from the semantic indexing dataset with a pool of 346 concepts, into a novel supervised multitask $ell !_p$ -norm dictionary learning framework. Intensive experimental results on TRECVID multimedia event detection dataset demonstrate the efficacy of our proposed technique.