HARadNet: Anchor-free target detection for radar point clouds using hierarchical attention and multi-task learning

Target localization and classification from radar point clouds is a challenging task due to the inherently sparse nature of the data with highly non-uniform target distribution. This work presents HARadNet, a novel attention based anchor free target detection and classification network architecture in a multi-task learning framework for radar point clouds data. A direction field vector is used as motion modality to achieve attention inside the network. The attention operates at different hierarchy of the feature abstraction layer with each point sampled according to a conditional direction field vector, allowing the network to exploit and learn a joint feature representation and correlation to its neighborhood. This leads to a significant improvement in the performance of the classification. Additionally, a parameter-free target localization is proposed using Bayesian sampling conditioned on a pre-trained direction field vector. The extensive evaluation on a public radar dataset shows an substantial increase in localization and classification performance.

Subjects

Multi-task learning

Radar detection

Scene understanding

DDC Class

621.3: Electrical Engineering, Electronic Engineering

006.3: Artificial Intelligence

Lizenz

https://creativecommons.org/licenses/by/4.0/

Publication version

publishedVersion

Name

1-s2.0-S2666827022000147-main.pdf

Type

Main Article

Size

1.58 MB

Format

Adobe PDF

Options

HARadNet: Anchor-free target detection for radar point clouds using hierarchical attention and multi-task learning