Temporal LiDAR Frame Prediction for Autonomous Driving
arXiv:2012.09409
Abstract
Anticipating the future in a dynamic scene is critical for many fields such as autonomous driving and robotics. In this paper we propose a class of novel neural network architectures to predict future LiDAR frames given previous ones. Since the ground truth in this application is simply the next frame in the sequence, we can train our models in a self-supervised fashion. Our proposed architectures are based on FlowNet3D and Dynamic Graph CNN. We use Chamfer Distance (CD) and Earth Mover's Distance (EMD) as loss functions and evaluation metrics. We train and evaluate our models using the newly released nuScenes dataset, and characterize their performance and complexity with several baselines. Compared to directly using FlowNet3D, our proposed architectures achieve CD and EMD nearly an order of magnitude lower. In addition, we show that our predictions generate reasonable scene flow approximations without using any labelled supervision.
arXiv commentIn 3DV 2020
References in corpus (6)
- Decoupled Weight Decay Regularization
- An Empirical Evaluation of Generic Convolutional and Recurrent Networks for Sequence Modeling
- Learning Representations and Generative Models for 3D Point Clouds
- Professor Forcing: A New Algorithm for Training Recurrent Networks
- PointRNN: Point Recurrent Neural Network for Moving Point Cloud Processing
- PointPWC-Net: A Coarse-to-Fine Network for Supervised and Self-Supervised Scene Flow Estimation on 3D Point Clouds