2026 / ECCV DriveX Workshop
Ashkan Yousefi Zadeh, Zishuo Zhu, Xiaomeng Li, Andry Rakotonirainy, Sebastien Glaser, Ronald Schroeter, Patricia Delhomme, and Zahra Mehraban.
Read more
2026 / Advanced Engineering Informatics, under review
Ashkan Yousefi Zadeh, Xiaomeng Li, Andry Rakotonirainy, Ronald Schroeter, Sebastien Glaser, and Zishuo Zhu.
Read more
2026 / AHFE IHIET
Andry Rakotonirainy, Ashkan Yousefi Zadeh, Zishuo Zhu, Djamel Benrachou, Mohammed Elhenawy, Sebastien Glaser, Xiaomeng Li, Ronald Schroeter, Melaine Gouillou, and Patricia Delhomme.
Read more
2025 / XAI Conference
Ashkan Yousefi Zadeh, Xiaomeng Li, Andry Rakotonirainy, Ronald Schroeter, and Sebastien Glaser.
Read more
2024 / IEEE T-ITS
A. Yousefi Zadeh, A. Jamali, R. Mallipeddi, and H. Khayyam.
Read more
2024 / EAAI
Z. Mehraban, A. Yousefi Zadeh, A. Jamali, R. Mallipeddi, and H. Khayyam.
Read moreMy thesis develops a model for generating human-centric explanations for automated vehicles. I connect Human-Centred Artificial Intelligence, Explainable AI, Natural Language Processing, and autonomous driving to improve transparency and trust in safety-critical systems.
My current applied AI work includes multimodal driving datasets, VLM generation and evaluation, NLP analysis of human explanations, and human-subject studies for trust and explainability evaluation.
Principal Supervisor
Human Behaviour, Experimental Design, Data Analysis.
Supervisor
Intelligent Transport Systems, AI, Human Factors.
Supervisor
Human-Machine Interaction, Road Safety, Automated Vehicle Design.
Supervisor
Intelligent Transport Systems, Automated Vehicle Design, Safety Deployment.
ECCV 2026 / DriveX Workshop
Presented our NARRATE work with Tina Mehraban and discussed human-centred explanations for automated driving with the computer vision and automated mobility research community.
Open project
Perception, scene understanding, sensor evidence, and driving context.
Language generation, linguistic building blocks, and structured explanation content.
Explanation design grounded in user needs, transparency, interpretability, and trust calibration.
Engadget / automated-vehicles / Oct 10, 2026
Headline tracked from Engadget as a recent automated vehicle signal.
WIRED / automated-vehicles / Oct 9, 2026
Headline tracked from WIRED as a recent automated vehicle signal.
TechCrunch / automated-vehicles / Oct 9, 2026
Headline tracked from TechCrunch as a recent automated vehicle signal.
WIRED / ai / Oct 1, 2026
Headline tracked from WIRED as a recent AI signal.
The Guardian / ai / Oct 1, 2026
Headline tracked from The Guardian as a recent AI signal.
WIRED / ai / Sep 30, 2026
Headline tracked from WIRED as a recent AI signal.
IEEE Transactions on Intelligent Transportation Systems / journal-rss / Jul 13, 2026
The increasing complexity of automated urban environments demands accurate modeling of pedestrian crossing behavior. Existing approaches often rely on human annotations or vehicle-centric views, limiting their ability to predict pedestrian intentions in dynamic scenarios. In this paper, we propose an Intent-Trajectory Integrated Pedestrian Crossing Prediction framework (ITI-PCP) that explicitly couples high-level semantic intention inference with multimodal trajectory generation.
IEEE Transactions on Intelligent Transportation Systems / journal-rss / Sep 3, 2026
Environment perception based on multimodal data fusion is crucial for the development of autonomous driving (AD) technology. However, the existing fusion-based detection and tracking methods often struggle to efficiently utilize the multimodal complementary information. This limitation can significantly impact the safety of autonomous vehicles (AVs), especially in complex traffic scenarios and adverse weather conditions.
arXiv / arxiv-llm-vlm-av / Oct 8, 2026
Multimodal large language models have made remarkable progress in bridging vision and language, facilitating various perception tasks essential for human-machine interaction, robotics, and autonomous driving. However, existing MLLM-based perception methods predominantly rely on text-based coordinate representation, which suffers from excessive token overhead, or fixed-range quantization, which suffers from range and precision constraints, especially for 3D domains with unbounded spatial range and high localization accuracy requirements. To address these challenges, we propose a dynamic vector decoding method named DVD, which unifies the representation of 2D and 3D perception tasks.
My QUT runner-up entry, A Human-Centric eXplainable Automated Vehicle (XAV).
Highlights from QUT's behavioural change competition, in which I participated.
IEEE Consumer Technology Society.
IEEE Vehicular Technology Society.
Teaching Computing and Data for Engineers at Queensland University of Technology.
AVR3, formerly CARRS-Q, Queensland University of Technology.
Reviewing for IEEE Transactions on Intelligent Transportation Systems, IEEE Access, and Transportation Research Part F: Traffic Psychology and Behaviour (Elsevier).
Server-side AI project development, deployment, and debugging at HeyvaAI.
Prof. Ali Jamali was one of the rare people who genuinely changed the direction of my life. He guided me, believed in me, and opened doors I never imagined I could walk through.
His lessons, kindness, and legacy stay with me. Everything I achieve from this point on carries a part of his influence.