Anastasiia Breskina

Doctoral Researcher, Instituto Superior Técnico, Universidade de Lisboa

Research Interests: Temporally-Aware Behavioural Monitoring, Multimodal Video Analysis, Action Unit Recognition, Responsible AI, Dataset Construction, Online Assessment Integrity

Google Scholar / Scopus / ORCID / GitHub / LinkedIn / Email

---

About

Computer vision and machine learning researcher with active work in video-based behavioural analysis. Current research develops a privacy-aware multimodal framework for temporally-grounded behavioural analysis from video – modelling sequences of facial Action Units, gaze, and head pose over time rather than classifying isolated frames, with regulatory compliance (GDPR, EU AI Act) embedded in the architecture. Online assessment integrity serves as the primary validation domain; the methodology is designed for broader application across regulated environments.

---

Publications

Journal Articles

• Breskina A., Antoshchuk S. Emostudent: a dataset for complex student behaviour evaluation // Odes’kyi politechnichnyi universytet pratsi. – 2023. – Vol. 1, no. 67. – P. 54–59. DOI: 10.15276/opu.1.67.2023.07

• Breskina A. Development of an automated online proctoring system // Herald of Advanced Information Technology. – 2023. – Vol. 6, no. 2. – P. 163–173. DOI: 10.15276/hait.06.2023.11 Scopus

• Antoshchuk S., Breskina A. Evaluation metrics systematization for 2D human poses analysis models // Herald of Advanced Information Technology. – 2023. – Vol. 6, no. 1. – P. 26–38. DOI: 10.15276/hait.06.2023.2 Scopus

• Antoshchuk S., Breskina A. Human action analysis models in artificial intelligence based proctoring systems and dataset for them // Applied Aspects of Information Technology. – 2023. – Vol. 6, no. 2. – P. 190–200. DOI: 10.15276/aait.06.2023.14

Book Chapter

• Breskina A., Antoshchuk S. Models of information analysis in automated online proctoring systems // Lecture Notes on Data Engineering and Communications Technologies. – Cham: Springer, 2024. – Vol. 219. – P. 258–268. DOI: 10.1007/978-3-031-70959-3_13 Scopus

Conference Proceedings

• Antoshchuk S., Breskina A. EmoStudent: developing a dataset to analyse students’ emotional well-being // Proceedings of the 11th International Conference on Applied Innovations in IT (ICAIIT). – 2023. – Vol. 11, no. 2. – P. 75–80. DOI: 10.25673/112996 Scopus

• Breskina A., Antoshchuk S. Data handling in AI-based proctoring systems and strategies for enhancement // ISDMCI 2024: Intelligent Systems for Decision Making and Problems of Computational Intelligence, Khmelnitskyi / Ústí nad Labem, 20–23 June 2024. – P. 5–6. ResearchGate

• Breskina A. Human action analysis in automated online proctoring systems // Modern Information Technology 2023, Odesa, 17 May 2023. – P. 53–55. PDF

Ukrainian-Language Publication

• Breskina A. Розробка автоматизованої системи онлайн прокторингу [Development of an automated online proctoring system] // Інформатика та математичні методи в моделюванні. – 2023. – No. 13. – P. 180–189. DOI: 10.15276/imms.v13.no1-2.180 (Ukrainian-language parallel publication of the HAIT 2023 article, issued under doctoral programme requirements)

---

Education

Doctoral Researcher (in progress)

Instituto Superior Técnico, Universidade de Lisboa | Lisbon, Portugal | 2026 – present

• Programme: PDEIC – Computer Science and Engineering
• Research: Privacy-aware multimodal framework for temporally-grounded behavioural analysis

Doctoral Researcher (attended)

Odesa National Polytechnic University | Odesa, Ukraine | 2016 – 2023

• Specialisation: Computer Science – Machine Learning, Computer Vision, AI
• Doctoral coursework completed 2023; research continued through October 2025
• Degree not conferred
• 9 publications produced during this period; 4 Scopus-indexed

Master’s Degree

Odesa National Polytechnic University | Odesa, Ukraine | 2013 – 2015

• Specialisation: Information Control Systems and Technologies

Bachelor’s Degree

Odesa National Polytechnic University | Odesa, Ukraine | 2009 – 2013

• Specialisation: Computer Science (Information Control Systems)

---

Research Experience

Doctoral Researcher

Instituto Superior Técnico | Lisbon, Portugal | 2026 – present

• Privacy-aware multimodal framework for temporally-grounded behavioural analysis from video
• Temporal modelling of facial Action Unit sequences, gaze, and head pose
• Regulated high-stakes environments as target deployment context

Doctoral Researcher

Odesa National Polytechnic University | 2023 – 2025

• Systematic analysis of 33 commercial AI proctoring systems across three deployment categories
• Empirical evaluation of 3 deployed platforms – 25–30% true-positive detection rate documented (24 trials, 2024)
• Behavioural video dataset design and annotation pipeline construction
• Working proctoring system prototype: audio + keyboard activity monitoring modules
• 9 publications across journals, Springer book chapter, and conference proceedings

Teaching Assistant

Odesa National Polytechnic University | Odesa, Ukraine | 2016 – 2018

• Courses: IT Project Team Management; Information Data Analysis; Big Data Processing; Machine Learning
• Student levels: Bachelor, Master
• Independently developed Machine Learning curriculum; delivered laboratory and practical sessions

---

Professional Experience

Associate Software Engineer

Hewlett Packard Enterprise | Odesa, Ukraine | December 2015 – March 2017

HPE ALM (formerly Quality Center)

• REST API, SSO, Robot framework integration; UI enhancements; test automation; CI configuration
• Technologies: Java, JavaScript, Spring, JUnit, MSSQL, Oracle, Maven, Git, Jenkins

Junior Java Developer

Luxoft | Odesa, Ukraine | August 2013 – January 2015

Navigation digital maps conversion & preprocessing system

• Desktop SWT application for 3D map visualisation, verification, route calculation, compilation monitoring
• Technologies: Java, OSGi (Equinox), Ant, Maven, Perforce, SQLite, PostgreSQL, OpenGL, GLSL, JOGL, Jenkins, JIRA

Android 3D maps rendering (R&D)

• Android application for real-time 3D map rendering
• Technologies: Java, Android SDK, SQLite, OpenGL ES, GLSL, Git

Lab Assistant

Odesa National Polytechnic University / Luxoft – IT-Incubator Laboratory | Odesa, Ukraine | October 2011 – August 2013

CV-Car Project

• Kinect-driven car; coloured marker tracking system
• Technologies: C++, Qt, OpenCV

JOSM Plugin (Java OpenStreetMap Editor)

• Road width determination from satellite imagery
• Technologies: Java, Ant