Dynamic and Static Object Detection Considering Fusion Regions and Point-wise Features

TL;DR

This paper introduces a novel method for detecting static and dynamic objects in front of autonomous vehicles, integrating YOLOv3 and Bayesian filtering to enhance real-time environmental understanding.

Contribution

It presents a new fusion-based approach combining deep learning and Bayesian filtering for improved object detection and characterization in autonomous driving.

Findings

Effective detection of static and dynamic objects in real-time

Enhanced estimation of object position, velocity, and heading

Superior performance compared to existing methods

Abstract

Object detection is a critical problem for the safe interaction between autonomous vehicles and road users. Deep-learning methodologies allowed the development of object detection approaches with better performance. However, there is still the challenge to obtain more characteristics from the objects detected in real-time. The main reason is that more information from the environment's objects can improve the autonomous vehicle capacity to face different urban situations. This paper proposes a new approach to detect static and dynamic objects in front of an autonomous vehicle. Our approach can also get other characteristics from the objects detected, like their position, velocity, and heading. We develop our proposal fusing results of the environment's interpretations achieved of YoloV3 and a Bayesian filter. To demonstrate our proposal's performance, we asses it through a benchmark dataset and real-world data obtained from an autonomous platform. We compared the results achieved with another approach.