Kaiser Hamid

I am a Ph.D. student at Texas Tech University and a Graduate Research Assistant in the Human-in-the-loop Advanced Cognitive Engineering (HiACE) Lab, advised by Dr. Nade Liang. I also collaborate with the Applied Vision Lab at Texas Tech University, led by Dr. Hamed Sari-Sarraf.

I am also pursuing an M.S. in Electrical & Computer Engineering at Texas Tech University. I received my B.Sc. degree from Bangladesh University of Engineering & Technology (BUET).

I am open to research collaborations and internship opportunities. If you find me a good fit, feel free to reach out!

CV  /  Email  /  Google Scholar  /  LinkedIn  /  GitHub

Research

I research computer vision and vision-language-action (VLA) models for autonomous driving, with a focus on robust perception, reliable decision-making, and safety-critical evaluation.

TAROPE: Temporal Aperture Rotary Position Encoding for Video LLMs

We introduce TAROPE, a training-free temporal positional encoding method that improves how Video LLMs represent temporal information under limited frame budgets without retraining the model.

CoRE: Weakly Supervised Coarse-to-Fine Risk Evidence Learning in Driving Videos

We introduce CoRE, a weakly supervised framework for learning human-perceived risk in egocentric driving videos by transforming clip-level subjective risk ratings into prediction-effect supervision for perceived-risk scoring, temporal risk-support localization, and tracked-object contribution estimation.

When2Talk: When Should a Proactive In-Car Agent Talk?

We study when a proactive in-car agent should communicate with passengers during autonomous driving. When2Talk compares event-triggered communication with context-sensitive communication that adapts when to speak based on driving-event priority and passenger activity.

ICR-Drive: Instruction Counterfactual Robustness for End-to-End Language-Driven Autonomous Driving

We introduce ICR-Drive, a controlled benchmark for measuring instruction robustness in language-conditioned driving by pairing identical CARLA routes and simulator seeds with systematic counterfactual instruction families spanning goal-preserving and goal-conflicting variants.

Predicting the Next Move: A Systematic Review of Driver Intention Modeling for Autonomous Vehicles

A systematic review of driver intention modeling for autonomous vehicles, summarizing datasets, features, modeling paradigms, prediction horizons, and evaluation practices.

FSDAM: Few-Shot Driving Attention Modeling via Vision-Language Coupling

We propose FSDAM, a few-shot driver attention modeling framework that couples vision-language supervision with driving perception to improve data efficiency and generalization.

A New Evaluation Metric for Takeover Maneuver Quality: Comparing Human Drivers with Autonomous Driving Agents

We introduce a new evaluation metric for takeover maneuver quality, enabling systematic comparison between human drivers and autonomous driving agents.

Assessing the Potential of Google Location History (GLH) Data for Travel Behavior Research in the Context of Developing Country

An empirical assessment of Google Location History data for travel behavior research, focusing on feasibility, limitations, and practical considerations in developing-country contexts.

Professional Service

Conference Review: HFES'25, TRB'26

Source code from Jon Barron's website