FuSSI-Net: Fusion of Spatio-temporal Skeletons for Intention Prediction Network

TL;DR

This paper introduces FuSSI-Net, an end-to-end framework that combines skeletal and object detection features with various fusion strategies to predict pedestrian intentions accurately, even half a second before risky actions occur.

Contribution

The work presents a novel fusion-based pedestrian intention prediction network with new evaluation metrics and demonstrates improved accuracy in diverse lighting conditions.

Findings

Early fusion achieves AP of 0.89 in intention classification.

The framework predicts intentions up to 0.5 seconds before risky maneuvers.

Proposed metrics effectively evaluate intention prediction performance.

Abstract

Pedestrian intention recognition is very important to develop robust and safe autonomous driving (AD) and advanced driver assistance systems (ADAS) functionalities for urban driving. In this work, we develop an end-to-end pedestrian intention framework that performs well on day- and night- time scenarios. Our framework relies on objection detection bounding boxes combined with skeletal features of human pose. We study early, late, and combined (early and late) fusion mechanisms to exploit the skeletal features and reduce false positives as well to improve the intention prediction performance. The early fusion mechanism results in AP of 0.89 and precision/recall of 0.79/0.89 for pedestrian intention classification. Furthermore, we propose three new metrics to properly evaluate the pedestrian intention systems. Under these new evaluation metrics for the intention prediction, the proposed end-to-end network offers accurate pedestrian intention up to half a second ahead of the actual risky maneuver.