Understanding and Predicting The Attractiveness of Human Action Shot

TL;DR

This paper introduces the first systematic study of the attractiveness of human action shots, creating a new dataset and proposing a deep learning model to predict attractiveness based on subjective human judgments.

Contribution

It presents a new dataset of 8000 action shots, and a deep Siamese network with a hybrid loss for predicting attractiveness, addressing the subjective nature of attractiveness prediction.

Findings

Attractiveness of action shots is subjective but predictable.

The proposed model effectively predicts attractive human action shots.

Crowd-sourced ratings can reliably measure attractiveness.

Abstract

Selecting attractive photos from a human action shot sequence is quite challenging, because of the subjective nature of the "attractiveness", which is mainly a combined factor of human pose in action and the background. Prior works have actively studied high-level image attributes including interestingness, memorability, popularity, and aesthetics. However, none of them has ever studied the "attractiveness" of human action shot. In this paper, we present the first study of the "attractiveness" of human action shots by taking a systematic data-driven approach. Specifically, we create a new action-shot dataset composed of about 8000 high quality action-shot photos. We further conduct rich crowd-sourced human judge studies on Amazon Mechanical Turk(AMT) in terms of global attractiveness of a single photo, and relative attractiveness of a pair of photos. A deep Siamese network with a novel hybrid distribution matching loss was further proposed to fully exploit both types of ratings. Extensive experiments reveal that (1) the property of action shot attractiveness is subjective but predicable (2) our proposed method is both efficient and effective for predicting the attractive human action shots.