In this report, we introduce our solution to the Occupancy and Flow Prediction challenge in the Waymo Open Dataset Challenges at CVPR 2022, which ranks 1st on the leaderboard. We have developed a novel hierarchical spatial-temporal network featured with spatial-temporal encoders, a multi-scale aggregator enriched with latent variables, and a recursive hierarchical 3D decoder. We use multiple losses including focal loss and modified flow trace loss to efficiently guide the training process. Our method achieves a Flow-Grounded Occupancy AUC of 0.8389 and outperforms all the other teams on the leaderboard.
@article{arxiv.2206.10118,
title = {HOPE: Hierarchical Spatial-temporal Network for Occupancy Flow Prediction},
author = {Yihan Hu and Wenxin Shao and Bo Jiang and Jiajie Chen and Siqi Chai and Zhening Yang and Jingyu Qian and Helong Zhou and Qiang Liu},
journal= {arXiv preprint arXiv:2206.10118},
year = {2022}
}
Comments
1st Ranking Solution for the Occupancy and Flow Prediction of the Waymo Open Dataset Challenges 2022 (http://cvpr2022.wad.vision/)