Mohammad

Ph.D. Student

prof_pic.jpg

Hello, I’m Mohammad, a Ph.D. candidate in Computer Science at George Mason University, working in the RobotiXX lab.

Research Goal: I study how robots can learn to understand and act in the physical world with minimal human supervision. My work combines self-supervised learning from multiple sensor modalities, learned models of how the world responds to a robot’s actions, and real-time planning with those models on real hardware. I’m especially interested in representations and world models that aren’t tied to a single task, so the same ideas can carry across robots, environments, and embodiments.

news

Aug 29, 2026 Hydra: A Navigation World Action Model with Discrete Latent Planning and Continuous Flow-Matching Execution introduces a unified world model and planner that can perform planning in latent space is on arXiv.
Jul 8, 2026 HumAIN: Human-Aware Implicit Social Robot Navigation was accepted at IROS 2026.
Jun 30, 2026 Received a $10,000 Microsoft Azure Compute Grant through the Microsoft Research Accelerator Program to support the Hydra project.
Jun 15, 2026 Received the Best Student Presentation Award at the 1st Workshop on Long-term Deployments in the Wild (LoWi) at IEEE ICRA 2026 in Vienna, Austria, for the presentation of VertiFormer.
Feb 12, 2026 Zero-Shot Adaptation to Robot Structural Damage via Natural Language-Informed Kinodynamics Modeling introduces language-informed kinodynamics modeling for zero-shot adaptation to robot structural damage is on arXiv.
Feb 1, 2025 VertiFormer: A Data-Efficient Multi-Task Transformer for Off-Road Robot Mobility combines dynamics learning and planning for off-road navigation is on arXiv.
Dec 30, 2024 Social-LLaVA: Enhancing Robot Navigation through Human-Language Reasoning in Social Spaces introduces a vision-language model and dataset for human-like reasoning in socially aware robot navigation was accepted at IROS 2025.
Oct 1, 2024 M2P2: A Multi-Modal Passive Perception Dataset for Off-Road Mobility in Extreme Low-Light Conditions was accepted at IROS 2025.

selected publications

  1. hydra.gif
    Hydra: A Navigation World Action Model with Discrete Latent Planning and Continuous Flow-Matching Execution
    Mohammad Nazeri, Alexandyr Card, Samira Huber, and 6 more authors
    arXiv preprint arXiv:2608.28995, 2026
  2. zlik.gif
    Zero-Shot Adaptation to Robot Structural Damage via Natural Language-Informed Kinodynamics Modeling
    Anuj Pokhrel, Aniket Datar, Mohammad Nazeri, and 2 more authors
    arXiv preprint arXiv:2602.12385, 2026
  3. PMoE.gif
    Exploring Reflective Limitation of Behavior Cloning in Autonomous Vehicles
    Mohammad Nazeri, and Mahdi Bohlouli
    In 2021 IEEE International Conference on Data Mining (ICDM), 2021
  4. musohu.gif
    Toward Human-Like Social Robot Navigation: A Large-Scale, Multi-Modal, Social Human Navigation Dataset
    Duc M. Nguyen, Mohammad Nazeri, Amirreza Payandeh, and 2 more authors
    In 2023 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2023
  5. cahsor.gif
    CAHSOR: Competence-Aware High-Speed Off-Road Ground Navigation in SE (3)
    Anuj Pokhrel, Mohammad Nazeri, Aniket Datar, and 1 more author
    arXiv preprint arXiv:2402.07065, 2024
  6. vanp.gif
    VANP: Learning Where to See for Navigation with Self-Supervised Vision-Action Pre-Training
    Mohammad Nazeri, Junzhe Wang, Amirreza Payandeh, and 1 more author
    arXiv preprint arXiv:2403.08109, 2024