cv

Research Assistant

Computer Vision Laboratory, Stony Brook University, NY, USA

As a member of SBU's CVLab I have been working on two funded research projects under the supervision of Prof. Dimitris Samaras, Prof. Andrew Schwartz and Prof. Owen Rambow. I am focusing on modeling vision and language in its human and cognitive contexts in an attempt to improve the way machines 'think' aiming to align them with human's Theory of Mind (ToM).

• Computer Vision
• Natural Language Processing

Research Engineer

OrbDB / KTH Royal Institute of Technology, Stockholm, Sweden

Doing research and engineering on a Graph Machine Learning and Uncertainty Quantification project. OrbDB is a fresh start in database design, combining the latest advances in AI and Data Systems to build world’s first responsible knowledge graph database.

• Graph Neural Networks
• Graph Databases

Research Engineer

National Center for Scientific Research Demokritos, Athens, Greece

As a member of the Multimedia Analysis Group / Computational Intelligence Laboratory, I worked on several multimodal learning projects revolving around speech-to-text applications, as well as signal processing challenges focusing on mental health state detection.

• Speech Recognition
• Computer Vision

Research Assistant

Information Processing Laboratory, Athens University of Economics and Business, Athens, Greece

As a member of the NLP Group / Information Processing Laboratory, I wrote my thesis on automatic medical image diagnosis (Diagnostic Captioning) under the supervision of Assistant Prof. John Pavlopoulos and Prof. Ion Androutsopoulos. This work led to an ACL Findings publication, as well as capturing the 1st place in ImageCLEFmedical 2025 Automatic Image Captioning task.

• Speech Recognition
• Computer Vision

PhD

Stony Brook University

Computer Science

MSc

Athens University of Economics and Business

Computer Science

BSc/Exchange Semester

Stockholm University

Computer Science

BSc

Athens University of Economics and Business

Computer Science

A Data-Driven Guided Decoding Mechanism for Diagnostic Captioning

Association for Computational Linguistics (ACL)

We proposed a novel training framework for HTR and ASR designed to align the model’s output character frequency distributions with empirical ones, aiming to improve robustness to temporal and contextual intra-dataset shifts.

A Self-Supervised Learning Approach for Detecting Non-Psychotic Relapses Using Wearable-Based Digital Phenotyping

IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)

it concerned MagCIL’s approach for the 1st track of the '2nd e-Prevention challenge: Psychotic and Non-Psychotic Relapse Detection using Wearable-Based Digital Phenotyping'. First we present our approach for preprocessing and extracting features from the wearable’s raw data. We then propose a Transformer model for learning self-supervised representations from augmented features, trained on data from non-relapse days from each of the 9 patients of the challenge. We adopt two unsupervised methods for detecting relapse days as outliers. A separate unsupervised model is tuned for each patient using the validation data of the challenge.

Greek2MathTex: A Greek Speech-to-Text Framework for LaTeX Equations Generation

Proceedings of the 13th Hellenic Conference on Artificial Intelligence

We proposed an end-to-end system that harnesses the power of Automatic Speech Recognition (ASR) and Natural Language Processing (NLP) techniques to enable users to verbally dictate mathematical expressions and equations in natural language, which are subsequently converted into LaTeX format.

AUEB NLP group at ImageCLEFmedical caption 2023

Proceedings of the 2023 Conference and Labs of the Evaluation Forum (CLEF)

We presented the methods that the AUEB NLP Group experimented with during its participation in the 7th edition of the ImageCLEFmedical Caption sub-tasks, namely Concept Detection and Caption Prediction. We ranked 1st in Concept Detection and 3rd in Caption Prediction.