Exploring Diversity-Based Active Learning for 3D Object Detection in Autonomous Driving

Jinpeng Lin

doi:10.1109/TITS.2024.3463801

Exploring Diversity-Based Active Learning for 3D Object Detection in Autonomous Driving

2024-11-01

0

PRE

AI

DOI：10.1109/TITS.2024.3463801

Original

Original request for help

Save

Abstract

En 中文

3D object detection has recently received much attention due to its great potential in autonomous vehicle (AV). The success of deep learning based object detectors relies on the availability of large-scale annotated datasets, which is time-consuming and expensive to compile, especially for 3D bounding box annotation. In this work, we investigate diversity-based active learning (AL) as a potential solution to alleviate the annotation burden. Given limited annotation budget, only the most informative frames and objects are automatically selected for human to annotate. Technically, we take the advantage of the multimodal information provided in an AV dataset, and propose a novel acquisition function that enforces spatial and temporal diversity in the selected samples. We benchmark the proposed method against other AL strategies under realistic annotation cost measurements, where the realistic costs for annotating a frame and a 3D bounding box are both taken into consideration. We demonstrate the effectiveness of the proposed method on the nuScenes dataset and show that it outperforms existing AL strategies significantly.

Keywords:

Three-dimensional displays

Object detection

Costs

Annotations

Detectors

Uncertainty

Diversity reception

Feature extraction

Autonomous vehicles

Point cloud compression

Active learning

3D object detection

autonomous driving

Exploring Diversity-Based Active Learning for 3D Object Detection in Autonomous Driving

Abstract

Journal

Organization