Md Kamrujjaman Mobin
Software Engineer I  @  Samsung R&D  ·  B.Sc. Computer Science & Engineering, SUST

Md Kamrujjaman Mobin

Building on-device AI at Samsung R&D Bangladesh while pursuing research in computer vision, embodied AI, and multimodal learning.

5 Publications
3.83 GPA / 4.00
2 🥉 Kaggle Medals
Actively seeking research opportunities  ·  Computer Vision, Multimodal Learning & Embodied AI
Background

About Me

Research Interests

Computer Vision Embodied AI & VLN VLA Models Multimodal Learning Biomedical AI NLP Physical AI & Robotics Vision-Language Models Event-Based Vision Long-Horizon Video Understanding Spatial Reasoning Tool-Use LLM Agents Health AI LLM Safety & Robustness

IELTS Academic: 7.0 Nov 2025

L: 7.5 R: 7.0 S: 6.5 W: 6.0

Education

Shahjalal University of Science & Technology, Bangladesh
B.Sc. in Computer Science & Engineering
Feb 2020 – Jul 2025  ·  CGPA: 3.83 / 4.00

Thesis: AI-Driven Health System for Mortality Prediction Using ECGs with Transformers and Selective State-Space Models. Focused on deep sequence modeling, representation learning, and AI-driven healthcare analytics.

Thesis PDF

Technical Skills

Languages
PythonC++JavaKotlinJavaScriptC
ML / AI
PyTorchHuggingFaceScikit-learnOpenCVNumPyPandas
LLM / VLM
QLoRA Fine-tuningSFT / GRPOSelf-Consistency EnsemblingBatch InferenceLiteRT-LM
Research
Computer VisionNLPMultimodalEmbodied AIVLAEvent-Based VisionSignal Processing
Academic
Research WritingExperimental DesignScientific PresentationPeer Review
Industry

Professional Experience

Samsung R&D Institute Bangladesh

Software Engineer I  ·  Dhaka, Bangladesh

July 2025 – Present
Personal Health Agent (PHA) On-device LLM Agent

On-device Android (Kotlin) agent that unifies data from multiple health platforms (Samsung Health, MyFitnessPal, Strava, etc.) via Google Health Connect and an MCP-style tool architecture, answering user queries with personalized insights and cross-app correlations. Fine-tuned Gemma 4 2B-IT for tool calling and compact system prompts, and deployed it with LiteRT-LM using on-device GPU acceleration to substantially reduce inference time.

Brain Pulse App Launching 2027

Android app for longitudinal cognitive health monitoring via on-device ML. Multi-module Clean Architecture (Hilt DI, Room DB, WorkManager); passive pipelines for app usage, screen events, sleep, and call logs; on-device ONNX/TFLite inference over a 10+ biomarker feature extraction pipeline.

RSDS Mobile SDK Remote Sensor Data Service

Clinical-grade SDK collecting 40+ sensor modalities from smartphone & Wear OS (PPG, ECG, EDA, BIA, accelerometer, sleep stages, app usage). KMP multi-module architecture with compile-time DI, SQLCipher-encrypted storage, GZIP sync, and AES file encryption for HIPAA-compliant data handling.

Compass Watch App Galaxy Smartwatch 7 & 8

Wearable sensor calibration utility. Refactoring achieved 7.5% SAM score improvement, 20% CBO reduction, and 90%+ unit test coverage.

Applied R&D Samsung Health · 450M+ users

Research on non-invasive blood glucose estimation from PPG signals and multimodal Alzheimer's risk modeling targeting Samsung Health's global platform.

Academic

Publications

Full list on Google Scholar.

Technical Reports · 2026
EBMV @ ECCV 2026

Training-Free Representation and Prompting for Event-Based Multimodal Understanding

Mobin, M.K., Shamil, M.S., & Arefeen, M.A.

Technical Report · EventBench General Understanding Track, Workshop on Event-Based & Multimodal Vision

Training-free system rendering event streams as polarity-coded, temporally ordered image grids for a frozen VLM (Qwen3.5-9B) with task-specific bin selection, cropping, and prompting. Ranked 9th of 27 teams (62.28% avg. across 5 tasks).

2025
Canadian AI 2025

Cardioformer: Advancing AI in ECG Analysis with Multi-Granularity Patching and ResNet

Mobin, M.K., Islam, M.S., Al Barid, S., & Masum, M.

38th Canadian Conference on Artificial Intelligence · Calgary, Canada

COLING 2025

LuxVeri at GenAI Detection Task 1: Inverse Perplexity-Weighted Ensemble for Robust Detection of AI-Generated Text across English and Multilingual Contexts

Mobin, M.K. & Islam, M.S.

31st International Conference on Computational Linguistics · Dubai, UAE

COLING 2025

LuxVeri at GenAI Detection Task 3: Cross-Domain Detection of AI-Generated Text Using Inverse Perplexity-Weighted Ensembles

Mobin, M.K. & Islam, M.S.

31st International Conference on Computational Linguistics · Dubai, UAE

2024
CinC 2024

Transforming ECG Images to Waveforms with UNet-LSTM and Multilabel Classification Using ResNet-LSTM: A Multimodal Approach

Mobin, M.K.

51st Computing in Cardiology · Karlsruhe, Germany

ICCIT 2024

CricPose360: A Comprehensive Dataset for Real-Time Pose-Enhanced Cricket Shot Classification Using YOLOv8

Mobin, M.K., Al Barid, S., & Rahman, M.S.

27th International Conference on Computer and Information Technology · Bangladesh

Academic Work

Research Projects

Active and ongoing research in embodied AI, neuromorphic computing, long-horizon video understanding, and multimodal & spatial reasoning.

Active · Jan 2026 Neuromorphic Vision · Robotics · Long-Horizon Video

Event-Based Vision & ANN–SNN Hybrid Perception

Investigating event-stream representations and temporal models for long-horizon egocentric video understanding on the newly released EventKitchen dataset (ECCV 2026). Earlier, surveyed 10+ papers (EventFly CVPR 2025, HsVT, SSM-based models) and ran experiments on N-EPIC Kitchens: voxel-grid generation, E2VID reconstruction, SSM+SNN temporal modeling. Proposing spike-native state-space models and noise-aware ANN-SNN hybrids.

2026 Embodied AI · ICLR 2025 Replication

GSA-VLN: General Scene Adaptation for Vision-Language Navigation

Replicated GR-DUET (ICLR 2025) on R2R/Matterport3D. Semantic-Aware Navigation with InfoNCE contrastive loss achieved +1.4% SR, +1.6% SPL. Experience Replay for continual learning: +0.7% late-episode SR.

2026 VLM Fine-tuning · EvalAI

MindCube Spatial Reasoning Challenge

Fine-tuned Qwen2.5-VL-3B-Instruct via QLoRA on MindCube using cognitive-map + chain-of-thought outputs. Built a 3-run weighted ensemble combining a custom SFT model and GRPO-trained RL model with self-consistency sampling. Structured outputs force explicit spatial reasoning before prediction; now extending this into a full research study.

Completed · 9th / 27 ECCV 2026 Workshop · EBMV

Event-Based Multimodal Understanding Challenge

General Understanding Track at ECCV 2026 Workshop on Event-Based & Multimodal Vision. Developed event-stream MLLM solutions for spatiotemporal reasoning and scene understanding. Training-free approach with a frozen Qwen3.5-9B ranked 9th of 27 teams (62.28% avg. across 5 tasks).

Ongoing LLM Agents · Tool Use

Serialization Bottleneck in Tool-Use Agents

Measuring how text-serialized intermediate representations (polygons, graphs, tables, ASTs) degrade compositional LLM agent pipelines. Evaluating runtime-managed object handles with deterministic property summaries as mitigation.

🥉 Bronze · Top 6.3% Kaggle · Computer Vision

Image Matching 2024 — Kaggle

Multi-model 3D reconstruction pipeline integrating COLMAP with deep feature matchers (LightGlue, SuperGlue, DKM, LoFTR).

🥉 Bronze · Top 5.7% Kaggle · Medical Imaging

ISIC 2024: Skin Cancer Detection — Kaggle

CNN-based multimodal models combining image and tabular data with advanced augmentation techniques for melanoma classification.

Open Source

Projects

Personal and open-source projects across AI, full-stack, and embedded systems.

MCP · LLM · Python

MCP Powered Weather AI

Conversational weather app with a custom Model Context Protocol (MCP) server and Gemini LLM. Supports standalone server and integrated client-server mode for terminal interactions.

LangChain · RAG · FastAPI

AskProfessor — RAG Conversational AI

Lightweight PDF Q&A using open-source LLMs in a Retrieval-Augmented Generation framework. LangChain orchestration, ChromaDB vector store, SQLite chat history, scalable FastAPI backend.

QLoRA · HuggingFace · LoRA

LLM Fine-Tuning

Fine-tuned LLaMA 3.1, Mistral, InstructGPT, and DeepSeek AI via QLoRA with n-bit quantization and prompt engineering. Optimized with LoRA adapters and mixed-precision training.

Node.js · MySQL · Tailwind

Inventory Management System

Full-stack web app for real-time inventory, sales, and order tracking with responsive UI. Complete lifecycle — MySQL schema, Express.js server-side logic, EJS templating.

Node.js · REST API · MySQL

E-commerce Website with API

Modular e-commerce platform with RESTful APIs for product management and supplier delivery. Integrated real-time order processing and buyer/seller authentication.

Java · MySQL · OOP

Virtual Classroom Using OOP

Google Classroom-inspired simulation built on core Java OOP principles. Features user login, assignment uploads, and submission tracking via a MySQL backend.

Arduino · C · ESP8266

IoT-Based Flood Detection System

IoT flood monitoring using Arduino, ultrasonic sensors, and ESP8266 for internet-based alerts. Real-time water level detection with automated email alerts on exceeding danger thresholds.

Connect

Get in Touch

Open to PhD & MSc opportunities, research collaborations, and interesting conversations.

Direct contact

Location Dhaka, Bangladesh