SMM Pasha

Syed Murtoza Mushrul Pasha

Researcher in AI security and medical AI · [Degree, Field], Chongqing University

[Email] GitHub Google Scholar LinkedIn CV (PDF)

About

I work on making machine learning systems trustworthy — especially defending them against backdoor attacks. My research spans AI security for IoT, edge, and federated learning systems, and medical AI for neuroimaging and diagnosis, where reliability and interpretability matter most.

I am currently seeking a PhD position in cybersecurity, with a proposed focus on AI-driven detection and mitigation of backdoor attacks in machine-learning-enabled IoT systems.

Backdoor attacks & defensesIoT & edge securityFederated learning securityLLM securityMedical imagingExplainable AI

Publications

Published

A Transfer Learning and Multi-Scale Feature Fusion Framework for Early-Stage Alzheimer's Disease Classification from MRI Images

SMM Pasha, [co-authors] · [Journal, Year]

Paper arXiv

Under review

MedBackdoorGuard: Detecting and Mitigating Backdoor Attacks in Deep Learning Models for MRI-Based Alzheimer's Disease and Mild Cognitive Impairment Diagnosis

SMM Pasha, [co-authors] · Submitted to Brain Sciences (MDPI)Under review

Code

SecureEdgeAgent: Security and Trust Aware Edge–Cloud Orchestration for Autonomous AI Agents

SMM Pasha, [co-authors] · Submitted to Internet of ThingsUnder review

Code

FedGuard-LLM: Cross-Round Backdoor Detection for Federated LLM Fine-Tuning

SMM Pasha, [co-authors] · Submitted to the 29th International Conference on Computer and Information Technology (ICCIT)Under review

Other research

Reinforcement-learning-based backdoor defense [title, venue, year]
Backdoor detection for IoT systems [title, venue, year]
MRLSA-Net — skin lesion segmentation [title, venue, year]
CausalLLM-AD — Alzheimer's disease [title, venue, year]
GALAD — drug–target interaction prediction [title, venue, year]

Research Experience

AI Security & Backdoor Defense · [Lab / Supervisor], Chongqing University
Designing detection and mitigation methods for backdoor attacks across medical imaging models (MedBackdoorGuard), edge–cloud AI agents (SecureEdgeAgent), and federated LLM fine-tuning (FedGuard-LLM).

[Years]

Medical AI & Explainable Diagnosis · [Lab / Supervisor]
Built a transfer-learning and multi-scale feature fusion model for early-stage Alzheimer's classification from MRI, with leakage-free evaluation and Grad-CAM interpretability; related work on skin lesion segmentation and drug–target interaction.

[Years]

Research Assistant — Automotive Software Quality
Assisted R&D of code quality control technologies for automotive electronic software, including data analysis and validation for industry-standard compliance.

[Years]

Education

[Degree], [Field]
Chongqing University, Chongqing, China

[Years]

Awards & Service

Star of International Students in Scientific and Technological Academic Innovation, Chongqing University — awarded to 9 international students university-wide. [Year]
Member, Chongqing Renewable Energy Society — presented on IoT security to 270+ members; supported Chinese–English translation at international meetings.
Student volunteer, Chongqing University — helped international students with daily life in China, including hospital visits and immigration procedures.

Curriculum Vitae

A full record of my education, publications, and research experience.

Download CV (PDF)

I welcome conversations about PhD opportunities and research collaborations in AI security — [Email].