MD Sameer Iqbal Chowdhury

MD Sameer Iqbal Chowdhury

Ph.D. Student, Computer Science, Texas State University

I study vision-language models, robotic manipulation, autonomous navigation, and intelligent sensing for reliable robotic systems. I am advised by Dr. Tsz-Chiu Au.

News

Research Interests

My current work explores how vision-language models can support robotics, especially failure detection and recovery in manipulation. I benchmark models ranging from edge-deployable systems such as Moondream2 and SmolVLM to larger reasoning models such as Qwen-2-VL 7B and Florence-2.

More broadly, I am interested in robotic perception, autonomous navigation and exploration, multimodal sensing, embedded intelligence, and resilient physical AI systems that can operate under noisy real-world conditions.

Publications

  1. J.1 MD Sameer Iqbal Chowdhury and Tsz-Chiu Au. 2026. Assessing Vision-Language Models for Failure Detection in Robotic Manipulation. IEEE Pulse. DOI: 10.1109/MPULS.2026.3659245. In press.
  2. S.1 MD Sameer Iqbal Chowdhury, Mikdam-Al-Maad Ronoue, Md Asaduzzaman, and Lafifa Jamal. 2024. Design of an IoT-enabled Scalable Closed Hydroponics System with Intelligent Parameter Control and Persistent Sensing Error Resilience. Smart Agricultural Technology. Under review. DOI: 10.2139/ssrn.5011949.
  3. C.1 Ipshita Ahmed Moon, Humayra Rashid, Safayat Hasan, MD Sameer Iqbal Chowdhury, Shifat-E-Arman, and Lafifa Jamal. 2024. Exploring the Efficacy of NAO Robot as a Language Instructor. Proceedings of the 10th International Conference on Robotics and Artificial Intelligence. DOI: 10.1145/3728985.3729006. Equal contribution.

Selected Projects

IoT-enabled hydroponics system setup

IoT-enabled Scalable Closed Hydroponics System

Designed and implemented an automated hydroponics system with pH, EC, CO2, humidity, and temperature monitoring. The system used Arduino-based nodes, a web dashboard, live monitoring, XGBoost-based control decisions, and a persistent sensing error-resilience algorithm.

The project received an ICT Innovation Fund grant from Bangladesh's ICT Division, where I served as Co-PI. Experiments showed a 78-288% improvement in fresh mass yield compared with conventional techniques.

Two degree-of-freedom single claw crawler robot

2-DOF Single Claw Crawler Bot using Q-learning

Built a two degree-of-freedom crawler robot and implemented Q-learning with epsilon-greedy exploration to optimize servo-angle sequences for maximum displacement.

Teaching

Doctoral Instructional Assistant, Texas State University, CS 2308: Foundations of Computer Science II, August 2025-present. I curate and grade assignments, proctor examinations, and provide office-hour support for two course sections with a combined cohort of 140+ students.

Leadership and Service

Education

Technical Skills

Languages: Python, C/C++, MATLAB. Frameworks: PyTorch, TensorFlow, YOLO, ROS/ROS2. Hardware: NVIDIA Jetson, Arduino, Raspberry Pi, sensors and actuators. Tools: Git, Docker, LaTeX, SolidWorks.

Honors and Awards

Contact

Email: vsj23@txstate.edu

Department of Computer Science, Texas State University. See my CV for the full academic record.