Single Image Action Recognition using Semantic Body Part Actions

TL;DR

This paper introduces a novel approach for single image action recognition that leverages semantic body part actions, combining deep learning and SVM fusion to improve accuracy over existing methods.

Contribution

It proposes a new method that models human actions through semantic body part actions using deep neural networks and SVM fusion, outperforming previous approaches.

Findings

Achieved 3.8% and 2.6% mAP improvements on PASCAL VOC 2012 and Stanford-40 datasets.

Demonstrated the effectiveness of semantic body part actions in action recognition.

Validated the approach with experiments on two benchmark datasets.

Abstract

In this paper, we propose a novel single image action recognition algorithm which is based on the idea of semantic body part actions. Unlike existing bottom up methods, we argue that the human action is a combination of meaningful body part actions. In detail, we divide human body into five parts: head, torso, arms, hands and legs. And for each of the body parts, we define several semantic body part actions, e.g., hand holding, hand waving. These semantic body part actions are strongly related to the body actions, e.g., writing, and jogging. Based on the idea, we propose a deep neural network based system: first, body parts are localized by a Semi-FCN network. Second, for each body parts, a Part Action Res-Net is used to predict semantic body part actions. And finally, we use SVM to fuse the body part actions and predict the entire body action. Experiments on two dataset: PASCAL VOC 2012 and Stanford-40 report mAP improvement from the state-of-the-art by 3.8% and 2.6% respectively.