Assistive Computer Vision
I am interested in perception systems that support blind and low-vision users in everyday environments. This includes pedestrian navigation, scene understanding, and reliable risk cues.
AI Researcher and Software Engineer
Diana Aleksieieva
AI researcher and software engineer pursuing an MS in Data Science and Engineering at The City College of New York. Former intern at The New York Times, with work spanning computer vision, assistive technology, and privacy-aware AI.
Research
Research interests
Visual Privacy
My work explores how models can detect and localize private or sensitive information in everyday images. The goal is responsible multimodal AI that helps users understand what visual data may reveal.
Multimodal Machine Learning
I study vision-language systems, open-vocabulary detection, grounding, segmentation, and verification. I care especially about how these models behave outside clean benchmark settings.
Urban and Public-Interest AI
I like projects that connect civic data, transportation, environmental systems, and public services. The strongest AI systems should make city life more understandable and humane.
Research
Publications and Research Writing
Research and reports that document work as it develops.
Reports2025Project report
WalkAware: Multitask Computer Vision for Pedestrian Accessibility
A technical report on combining segmentation, object detection, tracking, and structured risk logic to support safer pedestrian navigation for blind and low-vision users.
Open reportReports2025Project report
Flight110: Predicting JFK Airport Passenger Congestion
Developed a machine learning pipeline to forecast daily passenger throughput at JFK Airport two days ahead, supporting TSA staffing and operational planning.
Open reportFormal publications will be added as the research progresses.
Projects
Selected technical work.
A focused set of research, computer vision, civic technology, and data science projects.
Assistive Computer VisionMultitask real-time pipeline
WalkAware
How can real-time perception help blind and low-vision pedestrians interpret complex street scenes?
A multitask computer vision pipeline for pedestrian scene understanding, combining semantic segmentation, object detection, tracking, and rule-based risk analysis.
PythonPyTorchYOLOPIDNetSegFormerSAMOpenCV
SAMGrounding DINOOCRVision-language modelsPython
Civic Technology1.7M+ sanitation reports analyzed
Clean NYC Dashboard
NYC cleanliness data is spread across multiple public datasets, making it difficult to compare sanitation issues.
An interactive geospatial dashboard that analyzes more than 1.7 million NYC 311 sanitation reports alongside DSNY waste tonnage, litter basket locations, and population. Users can explore complaint density, recurring cleanliness issues, seasonal trends, and sanitation infrastructure across neighborhoods.
Next.jsLeafletReact-LeafletRechartsNYC Open Data APIsOpenStreetMap
Assistive TechnologyLast-Minute Door Guidance
DoorFront
Blind and low-vision users often need reliable, last-minute guidance to identify and navigate unfamiliar building entrances.
Improved its geospatial and automated image-collection pipelines, developed an administrative analytics dashboard with an annotation review and quality-control system.
FastAPIComputer visionOpenStreetMapNYC Open DataMongoDBGoogle Street View
SustainabilityBYTE Hacks 2025 Winner
GrukAI
People often struggle to identify how everyday waste should be disposed of, leading recyclable or compostable materials to end up in landfills.
Built a mobile-first AI platform that scans waste, identifies its material, recommends the correct disposal method, maps nearby recycling facilities, and encourages sustainable habits.
ReactGemini AIFirebaseFirestoreGoogle MapsTailwind CSS
Work Experience
2026
The New York Times
Software Engineering Intern
New York, NY
Worked with a professional engineering team to identify workflow problems, rapidly learn internal systems, and build practical tools that improved team processes.
Engineering workflowsInternal toolsProduct thinking
2025-present
Research Foundation of CUNY
Research Assistant
New York, NY
Developed and evaluated computer vision and vision-language systems for accessibility, including a visual privacy benchmark with 3,500+ annotated instances. Conducted controlled experiments, model evaluation, and failure analysis to improve visual privacy detection and accessible video description.
Computer visionEvaluationResearch systems
2025-present
Borough of Manhattan Community College
Supplemental Instruction Leader and Tutor
New York, NY
Mentored students in technical and research projects and led C++ and Discrete Structures workshops for 100+ students, emphasizing practical programming, AI literacy, and AI-assisted software development.
TeachingPythonDatabasesAlgorithms
Before software engineering
Earlier Design Experience
Graphic Designer and Team Lead
Ukraine and New York
Led a team of 8+ designers, managed client projects, and increased project output by 200%. Specialized in digital design, image composition, color representation, and visual media production.
Visual designProductionTeam leadership
Education
A path through data science, computing, and media technology.
GPA 4.0
City College of New York
MS in Data Science and Engineering
Machine learning, computer vision, data systems, and AI research.
GPA 4.0
Borough of Manhattan Community College
Associate degree coursework in Computer Science
Programming, algorithms, databases, web development, and systems.
GPA 3.7
Igor Sikorsky Kyiv Polytechnic Institute
B.S./M.S. Multimedia Technology and Engineering
An engineering and multimedia technology foundation that contributes to Diana's combination of visual design, technology, and human-centered problem solving.
Awards and Community
Columbia DivHacks 2025 — Best Use of Tavily MCP ServerBYTE Hacks 2025 — Planetary Protection WinnerTechnology Talent Pipeline participantCodePath TIP102 graduateGoIT Full-Stack Development graduateComputer Science Tutor and SI LeaderWomen in Data Science participantAccessibility and civic technology researcher
Beyond Work
Hiking, travel, and seeing new places.
