Naimul Haque

Naimul Haque

Entrepreneur · PhD Student · Computer Vision & AI Researcher
PhD Student & Graduate Research Assistant
School of Electrical Engineering & Computer Science
Louisiana State University (LSU), Baton Rouge, LA
Founder @ Gradmate.ai
naimul@gradmate.ai nhaque4@lsu.edu Baton Rouge, LA 70803

About

I am a PhD student and Graduate Research Assistant at Louisiana State University (LSU), where my research sits at the intersection of computer vision, large language models, and multimodal AI. My current doctoral work investigates catastrophic forgetting in sequential fine-tuning of LLMs, and the application of LLM-based multi-agent systems to smart-grid management (AI Copilot for GE Vernova GridOS).

In parallel, I am the founder of Gradmate.ai — an AI-powered platform that streamlines graduate school applications through LLM orchestration, program matching, SOP generation, and IELTS practice. Prior to LSU, I was a Graduate Research Assistant at Alfred University, where I won 1st place (tied) in the Advizex AI Innovation Challenge with an AI sales bot projected to improve ROI by 25%.

My publication record spans IEEE conferences, ICPR, and arXiv, with work in facial attribute captioning, sign-language recognition, low-resolution license plate recognition, and toxic content classification. I have also released open datasets and fine-tuned models on HuggingFace that are actively used by the community.

Research Interests

Computer Vision Face Analysis & Synthesis Large Language Models Catastrophic Forgetting Multimodal Systems Multi-Agent AI Image Captioning GAN & Diffusion Models Natural Language Processing Deep Learning AI for Energy Systems AI Ethics

Selected Projects

W-Space Refined Face Swapping Pipeline
Computer Vision · Generative AI
Neural Facial Identity Disentanglement via W-Space Refined StyleGAN Swapping

A face-swapping pipeline that disentangles facial identity from expression and appearance in StyleGAN's W-space. Introduces DeepFace neutral-emotion filtering to source only neutral-expression StyleGAN faces, paired with InsightFace gender validation for controlled attribute transfer. Key contribution: W-Space Refined blending outperforms Hybrid fusion (ID↔Src cosine similarity 0.359 vs. 0.232), enabled by convex-hull face masks from 106 landmarks, fixed Poisson blending with ROI boundary clamping, and pre-blend colour correction for perceptually natural composites.

StyleGAN DeepFace InsightFace Poisson Blending W-Space Face Disentanglement
FaceGemma Portrait Captioning Results
Computer Vision · Image Captioning · ICPR
FaceGemma: Attribute-Aware Portrait Image Captioning

FaceGemma enhances a Gemma-based vision-language model with an explicit facial attribute encoder, producing rich, identity-aware captions for portrait images. Trained on the custom FaceAttDB multilingual dataset (Zenodo), the model improves BLEU and CIDEr metrics over general-purpose captioners by conditioning generation on age, gender, emotion, and facial geometry.

Gemma VLM Facial Attributes Image Captioning FaceAttDB ICPR 2024
Bangla License Plate Recognition
Computer Vision · OCR · IEEE
Automatic Bangla License Plate Recognition for Low-Resolution Imagery

A two-stage deep-learning pipeline for recognising Bangla license plates under real-world low-resolution conditions. A YOLO-based detector localises the plate region, followed by a CNN-LSTM sequence model for character recognition. Evaluated on a custom dataset captured from surveillance cameras, achieving state-of-the-art recognition accuracy for Bangla script.

YOLOv8 CNN-LSTM OCR Bangla Script IEEE STI 2022
Portrait Colorization Results
Computer Vision · Generative · IEEE ICCIT
Grayscale Portrait Colorization via CNN and Pretrained VGG-Face Descriptors

An automatic portrait colorization framework that leverages identity-aware features extracted from a pretrained VGG-Face network as conditioning signals for a CNN colorization decoder. By anchoring colour prediction to facial identity features, the model produces perceptually plausible skin-tone and hair-colour outputs that are consistent across different images of the same subject.

CNN VGG-Face Colorization Portrait IEEE ICCIT 2019
Entrepreneurship · NLP · LLM Systems
Gradmate.ai — AI Copilot for Graduate School Applications

Founded and engineered Gradmate.ai, an end-to-end LLM-powered platform for graduate applicants. Core features include semantic program discovery (dense retrieval over university programme embeddings), multi-agent SOP/LOR generation via LangGraph, IELTS adaptive practice, and a community knowledge base. Stack: LangChain · LangGraph · Groq · DeepSeek · Next.js. Currently seeking seed investment to scale.

LangChain LangGraph Groq RAG Multi-Agent EdTech
PhD Thesis · LLM × Energy Systems
AI Copilot for GE Vernova GridOS: LLM-Driven Smart Grid Management

Doctoral thesis developing a multi-agent LLM system integrated with GE Vernova's GridOS platform for real-time power grid decision support. Agents handle fault diagnosis, load-balancing recommendations, and operator-alert summarisation using LangChain tool-calling over ADMS/AEMS APIs. Evaluated against operator response latency and decision-accuracy benchmarks.

LangChain Groq ADMS Power Systems Multi-Agent GE Vernova
Research · LLM · Continual Learning
Catastrophic Forgetting in LLMs: Sequential Fine-Tuning on NLP Benchmarks

Systematic empirical analysis of how LLMs degrade on previously learned NLP tasks after sequential domain fine-tuning. Benchmarks span GLUE and SuperGLUE tasks with controlled fine-tuning order experiments across Orca, Qwen, and other open-weight models. Explores elastic weight consolidation, LoRA merging, and replay strategies as mitigation techniques. Under review at ACL ARR 2024.

Continual Learning GLUE LoRA EWC ACL ARR 2024
Computer Vision · Human-Computer Interaction
BdSpell: Real-Time Bangla Finger-Spelling Recognition via YOLO

A real-time Bangla Sign Language finger-spelling system built on a custom-trained YOLOv8 object detector operating at 30+ FPS on commodity hardware. A curated dataset of 36 Bangla hand-gesture classes was collected under varied lighting and background conditions, achieving 94.3% mAP. Published at ICRIC 2023.

YOLOv8 Sign Language Real-Time Bangla NLP ICRIC 2023

Publications

2024
Under Review
Catastrophic Forgetting in LLMs: A Comparative Analysis Across Language Tasks
Naimul Haque
ACL ARR 2024
2024
FaceGemma: Enhancing Image Captioning with Facial Attributes for Portrait Images
Naimul Haque, Iffat Labiba, Sadia Akter
International Conference on Pattern Recognition (ICPR)
2023
BdSpell: A YOLO-based Real-time Finger Spelling System for Bangla Sign Language
Naimul Haque, Meraj Serker, Tariq Bin Bashar
ICRIC 2023 — International Conference on Recent Innovations in Computing
2022
IEEE
Bangla Toxic Comment Classification and Severity Measure Using Deep Learning
Naimul Haque, Md. Bodrul, Abdullah, Towfiq, Mehorab
ICRPSET 2022 — International Conference on Recent Progresses in Science, Engineering and Technology
2022
IEEE
Automatic Bangla License Plate Recognition System for Low-Resolution Images
Naimul Haque, Samira Islam, Rabeiya Afrose, Mohammad Shorif Uddin
IEEE STI 2022 — 4th International Conference on Sustainable Technologies for Industry 4.0
2019
IEEE
Grayscale Portrait Colorization using CNNs and Pretrained VGG-Face Descriptor
Naimul Haque, Samin Shahriar Tokey
IEEE ICCIT 2019 — 22nd International Conference on Computer and Information Technology
Datasets & Models
2023
FaceAttDB: A Multilingual Dataset for Facial Attribute Captioning
Naimul Haque, Abida Sultana
Zenodo — June 2023
2023
Bangla Toxic Comments Database
Naimul Haque, Badrul
HuggingFace Datasets — June 2023
2024
GlueOrca & GlueQwen: Fine-tuned Models for GLUE Benchmark Tasks
Naimul Haque
HuggingFace Model Repository

Experience

Mar 2025 –
Present
Founder & Entrepreneur
Gradmate.ai
Founded and lead development of an AI-powered graduate application platform.
  • LangChain & LangGraph for multi-agent orchestration
  • Groq and DeepSeek for low-latency LLM inference
  • RAG-based program discovery and SOP generation pipelines
Jan 2026 –
Present
Graduate Research Assistant & PhD Student
Louisiana State University (LSU)
Research on catastrophic forgetting in LLMs and AI for smart energy grids.
  • Sequential fine-tuning experiments on GLUE/SuperGLUE benchmarks
  • LLM copilot system for GE Vernova GridOS (thesis)
Jan 2024 –
Dec 2025
Graduate Research Assistant
Alfred University
AI and deep learning research; setup of Advanced Power Distribution Management Lab with GE's ADMS and AEMS systems. Teaching Assistant for Python for Power Electronics.
Aug 2023 –
Jan 2024
Assistant Professor
Uttara University
Undergraduate courses: Discrete Mathematics, Data Structures, Structured Programming.
Apr 2019 –
Aug 2023
Assistant Professor & Lecturer
Manarat International University & Cognition.AI Lab
Taught advanced AI courses and supervised undergraduate CV/NLP research.
  • Courses: AI, Computer Vision, Neural Networks, Pattern Recognition
  • Supervised capstone projects in face recognition and object detection

Education

2026 –
Present
Ph.D. — Electrical Engineering
Louisiana State University (LSU), Baton Rouge, LA
2024
M.Sc. — Electrical Engineering
Alfred University, Alfred, NY
M.Sc. — Computer Science
Jahangirnagar University, Bangladesh
2019
B.Sc. — Computer Science & Engineering
Ahsanullah University of Science and Technology (AUST), Bangladesh

Technical Skills

Languages
Python · C/C++ · MATLAB · JavaScript
CV Frameworks
OpenCV · YOLOv8 · CNNs · Detectron2 · DLIB
Generative / Face
StyleGAN · DeepFace · InsightFace · Diffusers
ML / DL
PyTorch · TensorFlow · Keras · Scikit-learn
NLP / LLM
HuggingFace · LangChain · LangGraph · Transformers
Inference
Groq · DeepSeek · Ollama · vLLM
MLOps / Tools
Docker · Git · Google Cloud · W&B · LaTeX
Power Systems
GE ADMS · AEMS · GridOS · Power Analytics

Awards & Recognition

Advizex AI Innovation Challenge — 1st Place (Tied) · 2024
Developed an AI sales bot projected to increase ROI by 25%. Featured in Alfred University News press release.
EPRI Certification — Machine Vision in the Electric Industry with Deep Learning · Sep 2024
22 PDHs in Power Industry AI applications.
Multiple Programming Competitions
1st and 2nd place finishes in university-level programming contests.