Auto4D: Learning to Label 4D Objects from Sequential Point Clouds

TL;DR

Auto4D introduces an automatic annotation pipeline that efficiently generates accurate 4D object labels from LiDAR point clouds, significantly reducing human effort and improving label quality through iterative refinement and trajectory smoothing.

Contribution

It presents a novel method for automatic 4D object labeling from LiDAR data by decomposing labels into size and motion, utilizing iterative refinement and trajectory smoothing.

Findings

25% reduction in human annotation effort

Improved label accuracy through trajectory smoothing

Effective on large-scale driving datasets

Abstract

In the past few years we have seen great advances in object perception (particularly in 4D space-time dimensions) thanks to deep learning methods. However, they typically rely on large amounts of high-quality labels to achieve good performance, which often require time-consuming and expensive work by human annotators. To address this we propose an automatic annotation pipeline that generates accurate object trajectories in 3D space (i.e., 4D labels) from LiDAR point clouds. The key idea is to decompose the 4D object label into two parts: the object size in 3D that's fixed through time for rigid objects, and the motion path describing the evolution of the object's pose through time. Instead of generating a series of labels in one shot, we adopt an iterative refinement process where online generated object detections are tracked through time as the initialization. Given the cheap but noisy input, our model produces higher quality 4D labels by re-estimating the object size and smoothing the motion path, where the improvement is achieved by exploiting aggregated observations and motion cues over the entire trajectory. We validate the proposed method on a large-scale driving dataset and show a 25% reduction of human annotation efforts. We also showcase the benefits of our approach in the annotator-in-the-loop setting.