Real-time Pose Estimation from Images for Multiple Humanoid Robots

TL;DR

This paper evaluates and adapts real-time deep learning pose estimation models for humanoid robots, introducing a lightweight model and a new dataset to enhance robot behavior in competitive environments.

Contribution

It proposes a lightweight, real-time pose estimation model tailored for humanoid robots and introduces the HumanoidRobotPose dataset for improved training and evaluation.

Findings

The model operates in real-time on humanoid robots.

The HumanoidRobotPose dataset improves pose estimation accuracy.

Enhanced robot behaviors enabled by the model.

Abstract

Pose estimation commonly refers to computer vision methods that recognize people's body postures in images or videos. With recent advancements in deep learning, we now have compelling models to tackle the problem in real-time. Since these models are usually designed for human images, one needs to adapt existing models to work on other creatures, including robots. This paper examines different state-of-the-art pose estimation models and proposes a lightweight model that can work in real-time on humanoid robots in the RoboCup Humanoid League environment. Additionally, we present a novel dataset called the HumanoidRobotPose dataset. The results of this work have the potential to enable many advanced behaviors for soccer-playing robots.