DriveSceneGen: Generating Diverse and Realistic Driving Scenarios From Scratch
- 20 June 2024
- journal article
- research article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Robotics and Automation Letters
- Vol. 9 (8) , 7007-7014
Abstract
Realistic and diverse traffic scenarios in large quantities are crucial for the development and validation of autonomous driving systems. However, owing to numerous difficulties in the data collection process and the reliance on intensive annotations, real-world datasets lack sufficient quantity and diversity to support the increasing demand for data. This work introduces DriveSceneGen , a data-driven driving scenario generation method that learns from the real-world driving dataset and generates entire dynamic driving scenarios from scratch. Experimental results on 5 k generated scenarios highlight that DriveSceneGen is able to generate novel driving scenarios that align with real-world data distributions with high fidelity and diversity. To the best of our knowledge, DriveSceneGen is the first method that generates novel driving scenarios involving both static map elements and dynamic traffic participants from scratch. Extensive experiments demonstrate that our two-stage method outperforms existing state-of-the-art map generation methods and trajectory simulation methods on their respective tasks.Keywords
All Related Versions
Funding Information
- National University of Singapore
- Yinson Green Technologies Pte Ltd.
- Moovita Pte Ltd.
- Nvidia
- Advanced Robotics Centre of the National University of Singapore on autonomous system technologies
This publication has 12 references indexed in Scilit:
- TrafficGen: Learning to Generate Diverse and Realistic Traffic Scenarios2023
- RePaint: Inpainting using Denoising Diffusion Probabilistic Models2022
- High-Resolution Image Synthesis with Latent Diffusion Models2022
- Lane-Level Street Map Extraction from Aerial Imagery2022
- Large Scale Interactive Motion Forecasting for Autonomous Driving : The Waymo Open Motion Dataset2021
- SceneGen: Learning to Generate Realistic Traffic Scenes2021
- HDMapGen: A Hierarchical Graph Generative Model of High Definition Maps2021
- SimNet: Learning Reactive Self-driving Simulations from Real-world Observations2021
- End-to-End Object Detection with Transformers2020
- U-Net: Convolutional Networks for Biomedical Image Segmentation2015