AA3DNet: Attention Augmented Real Time 3D Object Detection

TL;DR

This paper introduces AA3DNet, a real-time 3D object detection neural network using point cloud data, combining attention mechanisms and optimized training to achieve high accuracy and speed for autonomous vehicle perception.

Contribution

The paper presents a novel neural network architecture with attention modules and custom loss functions for real-time 3D object detection from point clouds, surpassing previous methods in accuracy and speed.

Findings

Achieves >30 FPS inference speed.

Outperforms previous state-of-the-art in average precision.

Demonstrates generalizability through ablation studies.

Abstract

In this work, we address the problem of 3D object detection from point cloud data in real time. For autonomous vehicles to work, it is very important for the perception component to detect the real world objects with both high accuracy and fast inference. We propose a novel neural network architecture along with the training and optimization details for detecting 3D objects using point cloud data. We present anchor design along with custom loss functions used in this work. A combination of spatial and channel wise attention module is used in this work. We use the Kitti 3D Birds Eye View dataset for benchmarking and validating our results. Our method surpasses previous state of the art in this domain both in terms of average precision and speed running at > 30 FPS. Finally, we present the ablation study to demonstrate that the performance of our network is generalizable. This makes it a feasible option to be deployed in real time applications like self driving cars.