MIND: Multi-modal Integrated PredictioN and Decision-making with Adaptive Interaction Modality Explorations

Tong Li^*†, Lu Zhang^*, Sikang Liu, Shaojie Shen

HKUST Aerial Robotics Group   
^*Equal Contributions ^†Corresponding Author

📃 Abstract

Navigating dense and dynamic environments poses a significant challenge for autonomous driving systems, owing to the intricate nature of multimodal interaction, wherein the actions of various traffic participants and the autonomous vehicle are complex and implicitly coupled. In this paper, we propose a novel framework, Multi-modal Integrated predictioN and Decision-making (MIND), which addresses the challenges by efficiently generating joint predictions and decisions covering multiple distinctive interaction modalities. Specifically, MIND leverages learning-based scenario predictions to obtain integrated predictions and decisions with social-consistent interaction modality and utilizes a modality-aware dynamic branching mechanism to generate scenario trees that efficiently capture the evolutions of distinctive interaction modalities with low variation of interaction uncertainty along the planning horizon. The scenario trees are seamlessly utilized by the contingency planning under interaction uncertainty to obtain clear and considerate maneuvers accounting for multi-modal evolutions. Comprehensive experimental results in the closed-loop simulation based on the real-world driving dataset showcase superior performance to other strong baselines under various driving contexts.

🔎 Quantitative Comparison of AIME

🔎 Qualitative Results On Argoverse 2

🛠️ Getting started

Create a new conda virtual environment

conda create -n mind python=3.10
conda activate mind

Install dependencies

pip install -r requirements.txt 

🕹️ Run a closed-loop simulation

python run_sim.py --config configs/demo_{1,2,3,4}.json

• The whole simulation takes about 10 minutes to finish. You are supposed to get the rendered simulation results saved in the outputs folder.

❤️ Acknowledgment

We would like to express sincere thanks to the authors of the following packages and tools:

• SIMPL
• ILQR