Danya Li

I am a Ph.D. student at the Intelligent Transport Systems (ITS) group at Technical University of Denmark (DTU), advised by Prof. Rico Krueger and Prof. Francisco C Pereira. My research focuses on machine learning for human behavior prediction in the context of autonomous driving.

My work spans virtual reality (VR)-based user studies, human behavior prediction, multimodal sensor fusion, and vision-language models — with the goal of making autonomous systems safer and more human-aware.

I obtained my M.Sc. from EPFL (Civil Engineering, minor in Data Science) at Switzerland and my B.Eng. from Tongji University (Transportation Engineering) at China. I was also a visiting researcher at TU Delft, Netherlands.

Research Interests

Pedestrian-AV Interaction Vision-Language Models Virtual Reality Spatio-temporal Modeling Multimodal Learning Transfer Learning

Danya Li

Ph.D. Student
Technical University of Denmark (DTU)
Copenhagen

danli@dtu.dk Google Scholar LinkedIn GitHub

Publications

Decoding Pedestrian Crossing Intentions from Egocentric Vision via Vision-Language Models

Danya Li, Xiang Su, Yan Feng, Rico Krueger

Best paper at ICRA workshop on Pedestrian Behavior Prediction, 2026

Paper

Eye Gaze-Informed and Context-Aware Pedestrian Trajectory Prediction in Shared Spaces with Automated Shuttles: A Virtual Reality Study

Danya Li, Yan Feng, Rico Krueger

Under review, 2025

Paper Code

Understanding and Predicting Human Behavior in Interactions with Autonomous Systems in Urban Environments: A Systematic Review, Challenges, and Opportunities

Danya Li, Wencan Mao, Francisco Pereira, Yu Xiao, Xiang Su, Rico Krueger

Transportation Research Part F: Traffic Psychology and Behaviour, 2025

Paper Website

TwoResNet: Two-Level Resolution Neural Network for Traffic Forecasting of Freeway Networks

Danya Li*, Semin Kwak*, Nikolas Geroliminis

IEEE Intelligent Transportation Systems Conference (ITSC), 2022 | Extended version: Scientific Reports, 2024

Paper (ITSC) Paper (Journal) Code

Using Machine Learning to Generate an Open-Access Cropland Map from Satellite Images Time Series in the Indian Himalayan Region

Danya Li, Joaquin Gajardo, Michele Volpi, Thijs Defraeye

Remote Sensing Applications: Society and Environment, 2023

Paper Code

Motion Style Transfer: Modular Low-Rank Adaptation for Deep Motion Forecasting

Parth Kothari, Danya Li, Yuejiang Liu, Alexandre Alahi

Conference on Robot Learning (CoRL), 2022

Paper Code

Experience

2023 – 2026

Ph.D. Researcher

Intelligent Transport Systems (ITS), DTU · Kongens Lyngby, Denmark

Thesis title: Modelling Pedestrian-Autonomous Vehicle Interactions Using Deep Learning, Artificial Intelligence, and Immersive Virtual Reality

2023 – 2026

Teaching Assistant

DTU · Kongens Lyngby, Denmark

Offered hands-on guidance on Model-based Machine Learning (probabilistic graphical models), Introduction to Business Analytics, and Advanced Business Analytics.

2024 – 2025

Visiting Researcher

Mobility in eXtended Reality Lab (MXR), TU Delft · Delft, Netherlands

Designed and conducted VR experiments on pedestrian-automated shuttles interactions.

2021 – 2022

Data Scientist Intern

Swiss Federal Laboratories for Materials Science and Technology (Empa) · St. Gallen, Switzerland

Developed an end-to-end machine learning pipeline for cropland detection from satellite image time series to support agricultural cold chain optimization.

2018

Transport Planning Intern

Shenzhen Urban Transport Planning Center · Shenzhen, China

Reviewed global bicycle system developement strategies and evaluated bicycle service level in the Nanshan district by field research.