Ehsan Ahmadi

Contact

Faculty of Science - Computing Science

Email: eahmadi@ualberta.ca

Overview

Area of Study / Keywords

Robotics Machine Learning Behavior modeling in Autonomous Driving Traffic Simulation

About

I am a Ph.D. student in Computing Science at the University of Alberta and a researcher at Noah’s Ark Lab, Huawei Technologies Canada. My research centers on behavior modeling in autonomous driving, with the goal of building traffic agents that simulate and predict real-world driving behavior. My recent work on RL-based fine-tuning for traffic simulation (RLFTSim) won 2nd place in the Waymo SimAgent Challenge 2025 and is accepted at CVPR 2026.

My research covers causal trajectory prediction and inter-agent interaction reasoning (CRiTIC, ICRA 2025), cooperative motion forecasting through vehicle-to-vehicle communication, and spatial reasoning with multimodal large language models (MLLMs). During my Master’s at the Sharif University of Technology, I was part of the CEDRA research group, where I developed a socially-aware SLAM algorithm and the ROS-based software framework for the Arash social robot — a companion robot for children with cancer. I also worked on robotic painting with the UR10 arm robot during a research visit at the University of Trieste.

Featured Publications

RLFTSim: Realistic and Controllable Multi-Agent Traffic Simulation via Reinforcement Learning Fine-Tuning

\textbfEhsan Ahmadi, Hunter Schofield, Behzad Khamidehi, Fazel Arasteh, Jinjun Shan, Lili Mou, Dongfeng Bai, Kasra Rezaee

CVPR. 2026 January;

CAPS: Context-Aware Priority Sampling for Enhanced Imitation Learning in Autonomous Driving

Arxiv. 2025 January;

Curb Your Attention: Causal Attention Gating for Robust Trajectory Prediction in Autonomous Driving

2025 IEEE INTERNATIONAL CONFERENCE ON ROBOTICS AND AUTOMATION, ICRA. 2025 January; 10.1109/ICRA55743.2025.11128367

Getting SMARTER for Motion Planning in Autonomous Driving Systems

2025 IEEE INTELLIGENT VEHICLES SYMPOSIUM, IV. 2025 January; 10.1109/IV64158.2025.11097471

AMEND: A Mixture of Experts Framework for Long-tailed Trajectory Prediction

Arxiv. 2024 January;