The ITSC2024 workshop, titled “Vision Language Model Based Human-Centered Autonomous Driving”, was organized by Tsukada lab’s Prof. Naren and Prof. Javanmardi, alongside Prof. Alexander Carballo (Gifu University), and Prof. Cristina Johannes (Johannes Kepler University Linz, Austria). The workshop explored the integration of vision language models (VLMs) into autonomous driving, with a focus on human factors in decision-making, safety, and interaction. The event featured five presentations, each covering unique aspects of human-centered autonomous driving technology.
Summary of Presentations:
- Automated Vehicle Driver Monitoring Dataset from Real-World Scenarios Integrating Naturalistic Insights in Objective Multi-Vehicle Safety Framework
- Presenter: Enrico Del Re, Research Assistant, Johannes Kepler University Linz
- Enrico discussed the use of real-world data to monitor drivers in automated vehicles, emphasizing the integration of naturalistic insights to form multi-vehicle safety frameworks.
- Large Language Models for Human-like Autonomous Driving Decision Making: A Survey
- Presenter: Yun Li, Doctoral Student, The University of Tokyo
- Yun Li surveyed the role of large language models (LLMs) in mimicking human-like decision-making for autonomous vehicles.
- Personalized Causal Factor Generalization for Subjective Risky Scene Understanding with Vision Transformer
- Presenter: Dr. Naren Bao, Assistant Professor, The University of Tokyo
- Dr. Naren explored the use of vision transformers for understanding subjective risky scenes, focusing on generalizing causal factors in a personalized way.
- Self-Training Systems for Protecting Wildlife
- Presenter: Dr. Stephany Berrio Perez, Research Associate, Australian Centre for Field Robotics
- Dr. Stephany presented a self-training system for protecting wildlife, contributing to ecological conservation through technology.
- Recent Activities Related to Driving Behavior and LLMs
- Presenter: Prof. Alexander Carballo, Gifu University
- Prof. Carballo discussed recent activities involving driving behavior and the application of LLMs for better modeling and predicting human driving patterns.
The workshop at ITSC2024 successfully brought together experts from diverse fields to discuss how vision language models and large language models can be integrated into autonomous driving. The presentations highlighted the importance of incorporating human factors into system design, aiming for responsive, adaptable, and socially acceptable autonomous vehicles.
Here are some photos of the ITSC2024 workshop. We had a peak attendance of 18 participants and received more than 10 questions throughout the session. A big thank you to Prof. Alex Carballo for his valuable assistance in the meeting room.