Introduction
I am a 5th-year PhD candidate in Information Systems at the University of Maryland, Baltimore County, a member of MPSC Lab and supervised by Prof. Nirmalya Roy. My research focuses on resource-aware real-time collaborative task execution for improved perception in heterogeneous autonomous systems.
My core research interests lie in Computer Vision, Edge Distributed Computing, and ROS-based autonomous systems. I have 4+ years of experience in designing scalable AI/ML solutions for real-time perception, multimodal sensor fusion, and task allocation in heterogeneous autonomous systems (UAVs & UGVs).
In my most recent work, I'm exploring the integration of Vision-Language Models (VLMs) into autonomous ground vehicles for enhanced situational awareness and adaptive decision-making in complex urban environments. I'm also working on improving 3D perception by fusing RGB, Lidar data to enhance detection accuracy and segmentation precision in dynamic, unstructured environments.
News
| 03/2026 | Paper titled "COHORT: Hybrid RL for Collaborative Large DNN Inference on Multi-Robot Systems Under Real-Time Constraints" accepted at IEEE WoWMoM 2026! |
| 03/2026 | Paper titled "CAViAR: Quality-Aware Vision-and-Radio Fusion for Relative Range Estimation among Collaborative Autonomous Agents" accepted at IEEE WoWMoM 2026! |
| 11/2025 | Paper titled "Imitation-Inspired Semantic-Guided Distillation for User-Conditioned Memorability Prediction" accepted at IEEE ICDM 2025. |
| 06/2025 | Paper titled "CoOpTex: Multimodal Cooperative Perception and Task Execution in Time-critical Distributed Autonomous Systems" accepted at DCOSS-IoT 2025. |
| 05/2025 | Paper titled "SkillNet: Human Actions Assessment via Human-AI Collaboration" accepted at ACM TOMM 2025. |
| 03/2024 | Paper titled "Arsfinetune: On-the-fly Tuning of Vision Models for Unmanned Ground Vehicles" accepted at DCOSS-IoT 2024. Congratulations to all co-authors! |
| 07/2023 | Paper titled "A Novel ROS2 QoS Policy-Enabled Synchronizing Middleware for Co-Simulation of Heterogeneous Multi-Robot Systems" accepted at ICCCN 2023. |
| 10/2023 | Paper titled "HeteroEdge: Addressing Asymmetry in Heterogeneous Collaborative Autonomous Systems" accepted at IEEE MASS 2023. |
Selected Publications
-
IEEE WoWMoM'26 et al. "COHORT: Hybrid RL for Collaborative Large DNN Inference on Multi-Robot Systems Under Real-Time Constraints," In Proc. of IEEE WoWMoM, 2026.
-
IEEE WoWMoM'26 Gaurav Shinde, Anuradha Ravi, Jared Lewis, Andre Harrison, Henry Gardiner, et al. "CAViAR: Quality-Aware Vision-and-Radio Fusion for Relative Range Estimation among Collaborative Autonomous Agents," In Proc. of IEEE WoWMoM, 2026.📊 Add figure:
images/caviar.png -
IEEE ICDM'25 I. Ghosh, , K. Jayarajah, and N. Roy, "Imitation-Inspired Semantic-Guided Distillation for User-Conditioned Memorability Prediction," In Proc. of the 25th IEEE International Conference on Data Mining, 2025.📊 Add figure:
images/icdm25.png -
DCOSS-IoT'25 et al. "CoOpTex: Multimodal Cooperative Perception and Task Execution in Time-critical Distributed Autonomous Systems," In Proc. of DCOSS-IoT, 2025.
-
ACM TOMM'25 I. Ghosh, Avijoy Chakma, et al. "SkillNet: Human Actions Assessment via Human-AI Collaboration," ACM Transactions on Multimedia Computing, Communications, and Applications, 2025.
-
Under Review et al. "SERN: Simulation-Enhanced Realistic Navigation for Multi-Agent Robotic Systems in Contested Environments."📊 Add figure:
images/sern.png -
Under Review et al. "MobHeteroCAS: Mobility-Aware DNN Task Scheduling in Heterogeneous Collaborative Autonomous Systems," ACM TAAS Journal, 2024.📊 Add figure:
images/mobhetero.png -
DCOSS-IoT'24 Ahmed, M., Hasan, Z., Faridee, A. Z. M., , Jayarajah, K., Purushotham, S., ... & Roy, N., "Arsfinetune: On-the-fly Tuning of Vision Models for Unmanned Ground Vehicles," Proc. of DCOSS-IoT, 2024.📊 Add figure:
images/arsfinetune.png -
ICCCN'23 Emon Dey, Mikolaj Walczak, et al. "A Novel ROS2 QoS Policy-Enabled Synchronizing Middleware for Co-Simulation of Heterogeneous Multi-Robot Systems," In Proc. of ICCCN, 2024.📊 Add figure:
images/icccn24.png -
IEEE MASS'23 , E. Dey, M. Devnath, I. Ghosh, N. Khan, J. Freeman, T. Gregory, N. Suri, K. Jayarajah, S. Ramamurthy, N. Roy, "HeteroEdge: Addressing Asymmetry in Heterogeneous Collaborative Autonomous Systems," In Proc. of IEEE MASS, 2023.
-
IEEE EICT'19 NB Noor, et al. "An Efficient Technique of Hemoglobin Level Screening Using Machine Learning Algorithms," IEEE EICT, 2019.📊 Add figure:
images/eict19.png -
BECITHCON'19 NB Noor, et al. "Comparative Study Between Decision Tree, SVM and KNN to Predict Anaemic Condition," IEEE BECITHCON, 2019.📊 Add figure:
images/becithcon19.png
Skills
Languages
ML / AI
Robotics & Hardware
Tools & Simulation
Experience
Graduate Research Assistant
Jan 2022 – PresentDeveloping frameworks for efficient UAV-UGV collaboration, optimizing task distribution and reducing communication overhead. Improving 3D perception by fusing RGB, LiDAR, and radar data. Integrating Vision-Language Models (VLMs) into autonomous ground vehicles for enhanced situational awareness.
Graduate Teaching Assistant
Jan 2022 – Dec 2023IS-603 (Decision-making Support System), IS-425 (Decision Support System), IS-733 (Data Mining).
Education
Doctor of Philosophy (PhD) in Information Systems
Jan 2022 – PresentPhD Topic: Resource-Aware Real-Time Collaborative Task Execution for Improved Perception In Heterogeneous Autonomous Systems
Advisor: Dr. Nirmalya Roy
Blog & News
-
2026 · Research Update
Two papers Accepted at IEEE WoWMoM 2026
First Authored COHORT and Co-authored CAViER have been accepted. We recently introduced COHORT, a ROS-based framework enabling heterogeneous robot teams to collaboratively execute large Vision-Language Models with real-time, energy-aware optimization. In parallel, our work on CAViAR presents a quality-aware multimodal fusion approach for robust relative distance estimation in GPS-denied environments
-
2025 · Research Update
Three Papers Published in 2025
Exciting year with publications at ICDM, DCOSS-IoT, and ACM TOMM covering distillation, cooperative perception, and human action assessment.
-
2025 · Project Spotlight
VLMs Meet Autonomous Ground Vehicles
Ongoing research integrating Vision-Language Models into UGVs for human-like reasoning in complex urban environments.