Aydin Ayanzadeh
CV

Computer vision · Multimodal learning · Medical AI

Aydin Ayanzadeh

Ph.D. student in Computer Science at the University of Maryland, Baltimore County, CORAL Lab, advised by Prof. Tim Oates.

I work on computer vision and multimodal learning for AI that is useful in clinics, in the field and in assistive technology: indoor navigation for people who are blind or have low vision, vision-language models for wildfire detection, and medical image segmentation under real data constraints.

Email Curriculum vitae Publications

Portrait of Aydin Ayanzadeh

Research

My current work at the CORAL Lab is accessible indoor navigation: turning a building's floor plan into a spatial knowledge graph that a language model can reason over, so blind and low-vision travelers get route guidance without per-building infrastructure.

I am the principal investigator of the human-subjects study behind it (UMBC IRB Protocol #1743, InVLM). Alongside, I work on vision-language models for early wildfire detection from satellite imagery, and on multi-task networks for medical images with imbalanced data.

Earlier, in Istanbul and Izmir: cell segmentation and tracking in label-free microscopy, and making deep models efficient through knowledge distillation, quantization for edge devices and graph representation learning.

From floor plan to route A floor plan with five rooms and a corridor. A graph of nodes and edges is laid over the rooms and doors, and a route is traced from the entrance, along the corridor, to the far room. start goal
Illustration: a floor plan becomes a graph of rooms and doors, and guidance is a path through it.
Perception & models
Computer Vision · Medical Image Analysis · Multimodal Learning · Large Language Models · Vision-Language Models
People & planet
Accessibility Technology · Environmental AI
Methods
Knowledge Distillation
Current focus
Accessibility AI · Environmental AI · Medical AI

Research threads, 2015–2026

Peer-reviewedPreprintActive years

Accessible navigation

Indoor navigation for blind and low-vision people: parsing floor plans into spatial knowledge graphs, agentic route planning with safety checks, and characterizing the on-device multimodal workload.

3 papers · 2025–2026

Wildfire and environmental AI

Vision-language models for early wildfire detection from satellite imagery, and auditable deep learning mechanisms for next-day wildfire spread prediction.

2 papers · 2026

Biomedical image analysis

Deep learning for cell segmentation in label-free phase-contrast microscopy, from still images to time-lapse wound-healing and cell-motility assays.

5 papers · 2019–2023

Efficient and representation learning

Knowledge distillation with clustered hint points, and residual variational graph autoencoders for link prediction.

3 papers · 2020–2023

Image processing and optimization

Earlier work on edge detection, transfer learning for fine-grained classification, and scaling matrices for trust-region methods.

4 papers · 2015–2019

One mark per paper, in the year it appeared. Select a mark to open the paper; the same 17 papers are in the publication library.

Toolkit

Languages, frameworks and platforms the work is built with.

  • PythonLanguages
  • MATLABLanguages
  • CLanguages
  • SQLLanguages
  • PyTorchFrameworks & libraries
  • TensorFlowFrameworks & libraries
  • KerasFrameworks & libraries
  • OpenCVFrameworks & libraries
  • Scikit-LearnFrameworks & libraries
  • PandasFrameworks & libraries
  • NumPyFrameworks & libraries
  • Git/GitHubTools & platforms
  • Google Cloud PlatformTools & platforms
  • AWSTools & platforms
  • LinuxTools & platforms
  • ImageJ/FijiTools & platforms

Selected publications

Recent and representative papers, each with its own page.

  1. 2026

  2. 2026

  3. 2026

  4. 2026

  5. 2025

  6. 2023

Browse all publications Google Scholar

News

  1. New preprint on arXiv: Modular Deep Learning Mechanisms for Auditable Next-Day Wildfire Spread PredictionPublication

  2. Paper accepted at IISWC 2026: Characterizing an On-Device Depth-Aware Multimodal Workload in Assistive Navigation for Blind and Low-Vision UsersPublication

  3. Paper at CVPR 2026 Workshop (GRAIL-V): LLM-Guided Agentic Floor Plan Parsing for Accessible Indoor Navigation of Blind and Low-Vision PeoplePublication

  4. Paper accepted at IGARSS 2026: WildfireVLM: AI-Powered Analysis for Early Wildfire Detection Using Satellite ImageryPublication

  5. Paper at IEEE BigData 2025: Floorplan2Guide: LLM-Guided Floorplan Parsing for BLV Indoor NavigationPublication

  6. M.S. Graduation & International Alumni ScholarshipAchievement

  7. Presenting at CMD-IT/ACM Richard Tapia ConferenceConference

  8. WildfireVLM: Environmental AI ResearchProject

  9. New Paper: Improved Autoencoder for Nuclei SegmentationResearch

  10. Presented at STARS Celebration ConferenceConference

  11. PURSUhInT Published in Expert Systems with ApplicationsPublication

  12. Vodafone FutureLab Research FellowshipAward

Research Projects

Current directions and selected published work.

Current Research

Active

LLM-Based Indoor Navigation System for Individuals with Blindness or Low Vision

2024 Under Review

Developing an innovative indoor navigation system using Large Language Models to assist individuals with visual impairments. The system combines computer vision, natural language processing, and spatial reasoning to provide real-time navigation guidance and environmental awareness.

Large Language Models Computer Vision Accessibility NLP
Collaborator: Prof. Tim Oates (UMBC)

Principal Investigator, UMBC IRB Protocol #1743 (InVLM) · Expedited, May 2025

View Project
Active

WildfireVLM: AI-Powered Analysis for Early Wildfire Detection and Risk Assessment

2024 Accepted

Developing a Vision-Language Model system for early wildfire detection using satellite imagery. The project focuses on creating an AI-powered solution that can analyze environmental data and provide early warning systems for wildfire prevention and management.

Vision-Language Models Satellite Imagery Environmental AI Risk Assessment
Collaborators: Prakhar Dixit, Sadia Kamal, Prof. Milton Halem (UMBC)
Accepted at: IEEE International Geoscience and Remote Sensing Symposium (IGARSS) 2026
View Project

Completed Research

PURSUhInT: Knowledge Distillation for Efficient Deep Learning

2023 Published

Developed a novel approach for knowledge distillation by identifying informative hint points based on layer clustering. This method significantly improves the efficiency of knowledge transfer from large teacher models to smaller student models.

Knowledge Distillation Deep Learning Model Compression Clustering
Published in: Expert Systems with Applications (Impact Factor: 8.5)
View Project

Deep Learning for Medical Image Segmentation

2020-2021 Published

Developed advanced deep learning pipelines for cell segmentation in label-free optical microscopy images. The work includes both traditional CNN approaches and novel architectures for improved accuracy in medical imaging applications.

Deep Learning Medical Imaging Computer Vision PyTorch
Published in: Turkish Journal of Electrical Engineering and Computer Sciences, IEEE SIU 2020
View Project

Graph Autoencoders for Representation Learning

2021 Published

Investigated residual connections in graph autoencoders for improved representation learning on graph-structured data. The work contributes to the field of graph neural networks and unsupervised learning.

Graph Neural Networks Autoencoders Representation Learning Unsupervised Learning
Published in: arXiv preprint, IEEE SIU 2020
View Project

Selected Projects

Technical implementations and research software, presented as a compact browsable collection.

Browse the full project hub

Deep Learning

Medical Image Segmentation

Multi-task deep learning platform for automated medical image analysis with advanced handling of imbalanced datasets.

PyTorch Computer Vision Medical AI
Read More
LLM App

LLM Indoor Navigation System

AI-powered navigation system using GPT-4 and computer vision to assist visually impaired individuals in indoor environments.

GPT-4 Computer Vision Accessibility
Read More
Optimization

Knowledge Distillation Framework

Novel hint-based knowledge distillation framework achieving 2.5x model compression with minimal accuracy loss.

PyTorch Model Compression Clustering
Read More
Multimodal AI

Vision-Language Medical Models

Custom CLIP-based architecture for medical imaging with zero-shot classification and report generation capabilities.

Transformers CLIP Multimodal
Read More
Graph Learning

Graph Autoencoder Framework

Advanced GNN implementation with residual connections for improved representation learning on graph-structured data.

PyTorch Geometric GNN Autoencoders
Read More
MLOps

ML Data Pipeline System

Scalable data processing pipeline handling 10TB+ daily with automated preprocessing and feature engineering.

Apache Spark Python MLOps
Read More
Environmental AI

WildfireVLM

Vision-Language Model for early wildfire detection and risk assessment from satellite imagery, accepted at IGARSS 2026.

Vision-Language Models Remote Sensing Environmental AI
Read More
Developer Tools

Claude Code Encyclopedia

Searchable reference for Claude Code, Codex CLI and GitHub CLI — commands, flags, settings and config paths in one page, with OS-aware examples.

Vanilla JS GitHub CLI Developer Tools
Read More
Assistive Tooling

ADHD Study Pack

A calm study desk for ADHD sessions: a focus dial, a week, wellbeing, and a shared table. English, Persian, Turkish, or Spanish. Google sign-in, or this browser only.

Accessibility Vanilla JS Firebase
Read More

Research experience

Research roles spanning accessible AI, medical imaging, computer vision, and efficient machine learning.

  1. Current – Present

    CORAL Lab, UMBC

    Researcher · Advisor: Tim Oates

    Baltimore, MD

    • Developing a multi-task network for segmenting and classifying medical images on imbalanced datasets
    • Leading InVLM accessible indoor navigation research with Prof. Tim Oates (project notes)
    • Principal Investigator, UMBC IRB Protocol #1743 (Expedited, approved May 2025)

  2. –

    Tubitak 1001 (Grant #119E578)

    Research Assistant · Supervisor: Devrim Unay

    Izmir, Turkey

    Project: Image processing and machine learning tools for phase-contrast optical microscopy time series

    • Analyzed cell morphology and movement in phase-contrast optical microscopy time series
    • Designed an auto-encoder that sped up training, improved tracking robustness, and reduced overfitting on segmentation and tracking

  3. –

    Arcelik Global Co.

    Researcher · Supervisor: Nazim Kemal Ure

    Istanbul, Turkey

    Project: Optimization of Multi-Task Network on Surveillance Cameras

    • Quantized a multi-task model for edge devices, raising inference from 2 fps to 18 fps

    Project: Model Compression for Efficient Video Processing on Edge Devices

    • Built a multi-task people-monitoring network for surveillance cameras on edge devices
    • Compressed the models with quantization, using Intel OpenVINO

  4. –

    Vodafone FutureLab

    Research Fellowship · Supervisor: Mehmet Basaran

    Istanbul, Turkey

    • Introduced Res-VGAE, a variational graph auto-encoder with residual connections
    • Proposed dynamic Word2Vec to examine social structure of Vodafone customers

  5. –

    SP4CING Lab

    Research Assistant · Co-Advisors: Behcet Ugur Toreyin and Devrim Unay

    Istanbul, Turkey

    • Introduced an auto-encoder with a modified ResNet-18 encoder

  6. –

    SiMiT Lab

    Researcher (Unpaid) · Supervisor: Hazim Kemal Ekenel

    Istanbul, Turkey

    Project: Kaggle Dog Breeds Identification with transfer learning

    • Fine-tuned ImageNet-pretrained models for the Kaggle dog-breed identification dataset

    Project: Google Cloud YouTube-8M Video Understanding Challenge

    • Examined deep neural networks with skip connections for the Kaggle video understanding challenge

Education

  1. Current – Present

    Ph.D. in Computer Science

    University of Maryland, Baltimore County

    GPA: 3.78/4.00

    Courses: Advanced Algorithms, Knowledge Graphs, Machine Learning, Computer Vision, Data Privacy

  2. –

    M.S. in Computer Science

    University of Maryland, Baltimore County

    GPA: 3.78/4.00

    Courses: Advanced Algorithms, Knowledge Graphs, Machine Learning, Computer Vision, Data Privacy

  3. –

    M.Sc. in Applied Informatics

    Istanbul Technical University, Turkey

    GPA: 3.75/4.00

    Courses: Image Processing, Applied Informatics in Structural Biology, Fuzzy Logic

  4. –

    B.Sc. in Computer Science

    University of Tabriz, Iran

    GPA: 3.11/4.00

Teaching & service

Graduate Teaching Assistant at UMBC since Spring 2022: systems, AI, data science, scripting, and intro programming.

Graduate Teaching Assistant

University of Maryland, Baltimore County · Spring 2022 – Present

  • CMSC 313Computer Organization and Assembly Language ProgrammingSpring 2025, Fall 2024, Spring 2024, Spring 2022
  • CMSC 104Problem Solving & ProgrammingSummer 2024
  • CMSC 433Scripting LanguagesSummer 2024
  • CMSC 671Principles of Artificial IntelligenceFall 2023
  • CMSC 304Social & Ethical Issues in ITSummer 2023
  • CMSC 691Introduction to Data ScienceSpring 2023, Fall 2022
  • CMSC 341Data StructuresSummer 2022
  • Supported undergraduate and graduate sections across systems, AI, data science, scripting, ethics, and intro programming
  • Ran labs and discussion support, held office hours, and graded assignments and exams

Journal reviewer

  • Expert Systems with Applications ElsevierJan 2024
  • Neurocomputing ElsevierNov 2024
  • Scientific Reports NatureMay 2024
  • Multimedia Systems Springer NatureMay 2024
  • International Journal of Machine Learning & Cybernetics Springer NatureJune 2024
  • Signal Processing: Image Communication ElsevierJune 2024

Mentoring

  • Undergraduate research in computer vision and machine learning
  • Technical guidance on programming projects and assignments
  • Career advising for research paths and graduate school
  • Project supervision for capstones and independent study in AI and vision

Awards & honors

  • International Alumni Endowed Scholarship

    University of Maryland, Baltimore County (UMBC), USA

    Prestigious scholarship recognizing outstanding academic achievement and international student contributions to the UMBC community.

  • Research Fellowship

    Vodafone FutureLab, Turkey

    Competitive fellowship for advanced research in telecommunications and technology innovation.

  • Top 1% Ranking

    Nationwide Universities Entrance Exam, Iran

    Achieved top 1% ranking among 500,000+ students in the highly competitive national university entrance examination.

Highlighted certificates

  • AI for Medicine (3-course specialization), deeplearning.ai on Coursera
  • Deep Learning (5-course specialization), deeplearning.ai on Coursera
  • Image and Video Processing, Duke University on Coursera

Contact

Interested in research collaborations, academic opportunities, or discussing ideas in AI and computer vision? I'd be happy to connect.

Reach me directly

University
aydina1@umbc.edu
Location
Baltimore, Maryland, USA
Lab
CORAL Lab, UMBC Computer Science

Send a message