Muhammad Ishfaq Hussain, Ph.D.

About

I specialize in vision-language modeling, medical imaging AI, and scalable training workflows. My current work involves fine-tuning and evaluating multimodal LLMs (e.g. MedGemma / PaliGemma variants) for visual question answering in radiology and broader scientific AI applications.

I enjoy translating research ideas into reproducible, performant systems while keeping an eye on interpretability and clinical usefulness.

Experience

Postdoctoral Research Fellow – Gwangju Institute of Science and Technology (GIST)

Aug 2025 – Present

• Artificial Intelligence (AI) Graduate School.
• Autonomous Driving (Machine Learning and Computer Vision for Robotics and Automation).
• Visual Large Language Models (VLLM’s).

Postdoctoral Research Fellow – KISTI (Large-Scale AI)

Jul 2024 – Aug 2025

• Fine-tuning multimodal LLMs (MedGemma / PaliGemma) on VQA-RAD & domain datasets.
• Medical VQA / Medical Image Segmentation.

Postdoctoral Research Fellow – Gwangju Institute of Science and Technology (GIST)

Sep 2023 – Jul 2024

• Machine Learning and Computer Vision for Robotics and Automation (Autonomous Driving).
• Visual Large Language Models (VLLM’s).

Ph.D. Researcher – GIST

2018 – 2023

• Thesis: predictive modeling through data fusion for autonomous driving vision systems.

Selected Projects

PMC-VQA Evaluation Suite

Benchmark & scripts for evaluating medical VQA models across multiple datasets with harmonized metrics.

PythonPyTorchEvaluation

Repo

MedGemma Fine-tuning Pipeline

Modular training harness with DeepSpeed & QLoRA, supporting mixed precision & gradient checkpointing.

DeepSpeedQLoRAMLOps

Repo

Autonomous Driving Sensor Fusion

Research prototypes for fusing multi-camera & LiDAR signals—improved object detection stability.

MultimodalCVFusion

Journal Articles

1. M. I. Hussain, L. M. Vinh, Z. Naz, U. Fatima, Y. Ko, and M. Jeon. SynthSWIRNet: Multi-spectral semantic fusion of synthetic SWIR with IR and RGB. Cluster Computing, Accepted (Nov 2025) Published 2026. DOI IF 4.1
2. Y. L. Jae, M. I. Hussain, K. Lee, H. S. Shim, S. Han, and Y. Donghun. Transfer learning-based super-resolution for high-precision medical imaging. IEEE Access, Accepted (July 2025).
3. M. I. Hussain, Z. Naz, M. A. Rafique, and M. Jeon. Advancing autonomous driving: DepthSense with radar and spatial attention. IEEE Sensors Journal 25(2):3698–3707, 2025. DOI IF 4.3
4. M. I. Hussain, M. A. Rafique, W.-G. Jung, B.-J. Kim, and M. Jeon. Segmentation and morphology computation of a spiky nanoparticle using the hourglass neural network. ACS Omega, 2023. DOI IF 4.2
5. M. I. Hussain, M. A. Rafique, J. Kim, M. Jeon, and W. Pedrycz. Artificial proprioceptive reflex warning using EMG in advanced driving assistance system. IEEE Trans. Neural Systems & Rehabilitation Engineering 31:1635–1644, 2023. DOI IF 4.9
6. M. I. Hussain, S. Azam, M. A. Rafique, A. M. Sheri, and M. Jeon. Drivable region estimation for self-driving vehicles using radar. IEEE Trans. Vehicular Technology 71(6):5971–5982, 2022. DOI IF 6.8
7. S. Azam, F. Munir, M. A. Rafique, A. M. Sheri, M. I. Hussain, and M. Jeon. N2C: Neural network controller design using behavioral cloning. IEEE Trans. Intelligent Transportation Systems 22(7):4744–4756, 2021. DOI IF 8.5
8. M. I. Hussain, M. A. Rafique, H. Jang, M. Jeon, and D. Shim. Camera-radar extrinsic calibration for autonomous driving. KIISE Transactions on Computing Practices 29(1):38–43, 2021.
9. Z. Khan, M. I. Hussain, N. Iltaf, J. Kim, and M. Jeon. Contextual recommender system for e-commerce applications. Applied Soft Computing 109:107552, 2021. DOI IF 8.263

Full list on Google Scholar →

Skills

Python PyTorch DeepSpeed QLoRA Transformers Vision-Language Models Medical Imaging Prompt Engineering Distributed Training Evaluation Pipelines Docker GitHub Actions

Contact

Email: ishfaq.gist@gmail.com

GitHub · LinkedIn · Scholar

Name Email Message

Send