Deep Reinforcement Learning for Active Human Pose Estimation

TL;DR

This paper introduces Pose-DRL, a deep reinforcement learning framework that actively selects viewpoints in space and time to improve 3D human pose estimation accuracy, outperforming traditional multi-view methods.

Contribution

It presents a novel active pose estimation architecture using deep reinforcement learning to select optimal viewpoints, including automatic stopping and transition strategies.

Findings

Significantly improves pose estimation accuracy over multi-view baselines.

Effective in complex scenes with multiple people.

Learns to select informative views in both spatial and temporal domains.

Abstract

Most 3d human pose estimation methods assume that input -- be it images of a scene collected from one or several viewpoints, or from a video -- is given. Consequently, they focus on estimates leveraging prior knowledge and measurement by fusing information spatially and/or temporally, whenever available. In this paper we address the problem of an active observer with freedom to move and explore the scene spatially -- in `time-freeze' mode -- and/or temporally, by selecting informative viewpoints that improve its estimation accuracy. Towards this end, we introduce Pose-DRL, a fully trainable deep reinforcement learning-based active pose estimation architecture which learns to select appropriate views, in space and time, to feed an underlying monocular pose estimator. We evaluate our model using single- and multi-target estimators with strong result in both settings. Our system further learns automatic stopping conditions in time and transition functions to the next temporal processing step in videos. In extensive experiments with the Panoptic multi-view setup, and for complex scenes containing multiple people, we show that our model learns to select viewpoints that yield significantly more accurate pose estimates compared to strong multi-view baselines.