VALOR
Vision and Language Oriented Representation:
Topics in Semantics, Safety, and Explainability in Intelligent Transportation
1st Workshop at ITSC 2024
Tuesday September 24 - Afternoon
About the Workshop
Foundation models, LLMs, and vision-language models have raised significant attention. Zero-shot learning provides the potential for intelligent applications to be created with minimal data collection. Language-based reasoning tools have made possible new modes of planning and control. This workshop aims to bridge the gap between the rapid advances occurring in these areas within the broader computer science community and the potential new technologies and safety applications that may be enabled for intelligent transportation systems (ITS). The intended workshop audience will include those within ITS who have an interest in learning about the newest vision-language developments or sharing their own findings and applications, as well as those within the vision-language community who have ideas to share about how these new capabilities might play a part in improving (or creating new) intelligent transportation systems, broadly defined.
Vision-language models and related systems that learn relationships between the two have recently provided a variety of interesting research directions within ITS. A few such examples include vision-language modeled end-to-end driving, novelty detection and explanation, data curation and active learning, scene segmentation, risk perception, control, and more. The rate of development in this field is quite rapid, and the goal of this workshop is to unite researchers from a variety of backgrounds, including, but not limited to, those with research insights into these models and representations, those with insights into ITS problems that such models may enable new solutions for, and those who have devised or applied new related systems of learning for ITS.
To this end, we welcome contributions with a strong focus on - but not limited to - the following topics:
VL and Dataset Curation, Data Labeling for ITS
VL in Machine Learning for ITS
VL in Explainability for ITS
VL in Multi-Agent Interactions for ITS
Representation Learning through VL for ITS
Foundation Models for ITS
VL Domain Adaptation for ITS
VL in Simulation for ITS
VL in Validation for ITS
Safety of ITS using VL
Workshop Speakers
Professor Manmohan Chandraker, University of California San Diego
Professor Kazuya Takeda, Nagoya University
Professor Abhinav Valada, University of Freiburg
Professor Ziran Wang, Purdue University
Dr. Neel Bhatt, University of Texas Austin
Dr. Trent Victor, Waymo
Workshop Presentations
Shounak Sural, Carnegie Mellon University
Hubert Padusinski, FZI Research Center for Information Technology
Malsha Mahawatta, University of Gothenburg
Workshop Schedule
Organizers
General Chairs
Professor Mohan Trivedi (University of California San Diego)
Professor Ross Greer (University of California Merced)
Walter Zimmer, MSc (Technical University of Munich)
Program Chairs:
Dr. Julie Stephany Berrio-Perez (The University of Sydney, Australia)
Dr. Nachiket Deo (Latitude)
Xingcheng Zhou (Technical University of Munich)
Student Chairs:
Akshay Gopalkrishnan (University of California San Diego)
Workshop Contact: rossgreer@ucmerced.edu