ActAR: Actor-Driven Pose Embeddings for Video Action Recognition

TL;DR

This paper introduces a novel method for human action recognition in infrared videos that automatically identifies key actors and key poses, improving recognition accuracy without requiring prior annotations.

Contribution

The paper presents a new actor-driven pose embedding approach that enhances infrared action recognition by automatically selecting key actors and poses, addressing challenges of texture ambiguity and multiple actors.

Findings

Achieves promising recognition performance on InfAR dataset.

Automatically identifies key actors without prior annotations.

Enhances action representations through pose filtering.

Abstract

Human action recognition (HAR) in videos is one of the core tasks of video understanding. Based on video sequences, the goal is to recognize actions performed by humans. While HAR has received much attention in the visible spectrum, action recognition in infrared videos is little studied. Accurate recognition of human actions in the infrared domain is a highly challenging task because of the redundant and indistinguishable texture features present in the sequence. Furthermore, in some cases, challenges arise from the irrelevant information induced by the presence of multiple active persons not contributing to the actual action of interest. Therefore, most existing methods consider a standard paradigm that does not take into account these challenges, which is in some part due to the ambiguous definition of the recognition task in some cases. In this paper, we propose a new method that simultaneously learns to recognize efficiently human actions in the infrared spectrum, while automatically identifying the key-actors performing the action without using any prior knowledge or explicit annotations. Our method is composed of three stages. In the first stage, optical flow-based key-actor identification is performed. Then for each key-actor, we estimate key-poses that will guide the frame selection process. A scale-invariant encoding process along with embedded pose filtering are performed in order to enhance the quality of action representations. Experimental results on InfAR dataset show that our proposed model achieves promising recognition performance and learns useful action representations.