VAD: Vectorized Scene Representation for Efficient Autonomous Driving

By Bo Jiang, Shaoyu Chen, Qing Xu, Bencheng Liao, Jiajie Chen, Helong Zhou, Qian Zhang, Wenyu Liu, Chang Huang, Xinggang Wang. IEEE International Conference on Computer Vision, 2023

729 citations recorded by the source corpus.

Autonomous driving requires a comprehensive understanding of the surrounding environment for reliable trajectory planning. Previous works rely on dense rasterized scene representation (e.g., agent occupancy and semantic map) to perform planning, which is computationally intensive and misses the instance-level structure information. In this paper, we propose VAD, an end-to-end vectorized paradigm for autonomous driving, which models the driving scene as a fully vectorized representation. The proposed vectorized paradigm has two significant advantages. On one hand, VAD exploits the vectorized agent motion and map elements as explicit instance-level planning constraints which effectively improves planning safety. On the other hand, VAD runs much faster than previous end-to-end planning methods by getting rid of computation-intensive rasterized representation and hand-designed post-processing steps. VAD achieves state-of-the-art end-to-end planning performance on the nuScenes dataset, outperforming the previous best method by a large margin. Our base model, VAD-Base, greatly reduces the average collision rate by 29.0% and runs 2.5× faster. Besides, a lightweight variant, VAD-Tiny, greatly improves the inference speed (up to 9.3×) while achieving comparable planning performance. We believe the excellent performance and the high efficiency of VAD are critical for the real-world deployment of an autonomous driving system. Code and models are available at https://github.com/hustvl/VAD for facilitating future research.

Subfields: Automotive Engineering, Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/aae2e7454945b904ca1d438d2ae4d4941ba11409

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Automotive Engineering

1. Junru Gu
2. Jiageng Mao
3. Hang Zhao
4. Yuxi Qian
5. Yue Wang

In Computer Vision and Pattern Recognition

1. Wenyu Liu
2. Qian Zhang
3. Xinggang Wang
4. Zhaoxiang Zhang
5. Bencheng Liao
6. Lue Fan
7. Hongxin Li
8. Chenxu Hu

Citations in the corpus

• DriveVLM: The Convergence of Autonomous Driving and Large Vision-Language Models (2024, methodology): “Following previous works [22, 24], we adopt Displacement Error (DE) and Collision Rate (CR) as metrics to evaluate models’ performance on the validation split.”
• Driving Into the Future: Multiview Visual Forecasting and Planning with World Model for Autonomous Driving (2023, methodology): “Based on plans from VAD [30], Drive-WM forecasts multiple plausible futures consistent with the initial observation.”
• NAVSIM: Data-Driven Non-Reactive Autonomous Vehicle Simulation and Benchmarking (2024, neutral): “Several submissions attempted to reimplement or extend prior work on nuScenes such as UniAD [21] and VAD [25], but were unable to outperform the TransFuser baseline by the challenge submission deadline, given the significant engineering challenge and compute requirements.”
• DiffusionDrive: Truncated Diffusion Model for End-to-End Autonomous Driving (2024, positive): “DiffusionDrive can integrate various existing perception modules used in previous end-to-end planners [8, 16, 20, 39] and take different sensor inputs.”
• VADv2: End-to-End Vectorized Autonomous Driving via Probabilistic Planning (2024, methodology): “In comparison to the previous result [23], VADv2 significantly improves Drive Score and Route Completion by 25.3 and 5.7 respectively, demonstrating the comprehensive driving ability of VADv2 in complex driving scenarios.”
• Vista: A Generalizable Driving World Model with High Fidelity and Versatile Controllability (2024, neutral): “To categorize actions into commands, we follow the established conventions in planning [56, 58, 64] and define the command of ego-vehicle as " turn right " or " turn left " when its final displacement exceeds 2 meters in the orthogonal direction relative to its initial heading.”
• GPT-Driver: Learning to Drive with GPT (2023, methodology): “We follow the general practice in prior works Hu et al. (2022; 2023); Jiang et al. (2023) and split the whole dataset into training, validation, and testing sets.”