MapBEVPrediction

This repository contains the official implementation of Accelerating Online Mapping and Behavior Prediction via Direct BEV Feature Attention published in ECCV 2024.

Note: This is an untested version. Most of the code are adopted from my previous work. Let me know if any (very likely) bugs appear and I will fix it ASAP. Will try to run this from scratch hopefully soon (still at internship so time is not guaranteed :( ). Thank you so much for your interest and support!!

Getting Started

• Environment Setup
• Prepare Dataset
• Mapping Train and Eval
• Merge Map and Trajectory Dataset
• Trajectory Train and Eval
• Visualization

Results

Mapping checkpoints are here. Trajectory prediction checkpoints are here.

Dataset

I have uploaded two sample datasets (complete) for MapTRv2 and MapTRv2 CL. They are around 500GB each. You can download them through AWS S3. They are located at

• AWS Bucket Name: s3://mapbevprediction
• Region: us-east-2

Dataset Structure is as follows:

mapbevprediction
├── maptrv2_bev/
│   ├── mini_val/
│   |   ├── data/
│   |   |   ├── scene-{scene_id}.pkl
│   ├── train/
│   ├── val/
├── maptrv2_cent_bev/

Catalog

• Visualization Code
• Code release
  • MapTR
  • MapTRv2
  • StreamMapNet
  • HiVT
  • DenseTNT
• Untested version released + Instructions
• Initialization

Citation

If you found this repository useful, please consider citing our work:

@Inproceedings{GuSongEtAl2024,
  author    = {Gu, Xunjiang and Song, Guanyu and Gilitschenski, Igor and Pavone, Marco and Ivanovic, Boris},
  title     = {Accelerating Online Mapping and Behavior Prediction via Direct BEV Feature Attention},
  booktitle = {European Conference on Computer Vision (ECCV)},
  year      = {2024}
}

This codebase is built using our prior work, if your found this helpful, please also consider citing:

@Inproceedings{GuSongEtAl2024,
  author    = {Gu, Xunjiang and Song, Guanyu and Gilitschenski, Igor and Pavone, Marco and Ivanovic, Boris},
  title     = {Producing and Leveraging Online Map Uncertainty in Trajectory Prediction},
  booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
  year      = {2024}
}

License

This repository is licensed under Apache 2.0.