Pavlos Andreadis
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Pavlos Andreadis has a total of 6 co-authors. Here are the top co-authors by number of shared publications:
- Dimitrios I. Diochnos
University of Oklahoma
1 shared publication - Jay C. Rothenberger
The University of Oklahoma
1 shared publication - Krystof Bezdek
School of Informatics, University of Edinburgh, Edinburgh, United Kingdom
1 shared publication - Marc Violides
School of Informatics, University of Edinburgh, Edinburgh, United Kingdom
1 shared publication - Patrick Kage
A:1:{s:5:"en_US";s:27:"The University of Edinburgh";
1 shared publication - Tanatip Timtong
School of Informatics, University of Edinburgh, Edinburgh, United Kingdom
1 shared publication
Latest Publications
- Top Cited Papers
- 17 March 2026
- Published by AI Access Foundation in Journal of Artificial Intelligence Research
- Vol. 85
A Review of Pseudo-Labeling for Computer Vision
Deep neural models have achieved state-of-the-art performance on a wide range of problems in computer science, especially in computer vision. However, deep neural networks often require large datasets of labeled samples to generalize effectively. An important area of active research is semi-supervised learning, which attempts to instead utilize large quantities of (easily acquired) unlabeled samples. One family of methods in this space is pseudo-labeling, a class of algorithms that use model outputs to assign labels to unlabeled samples which are then used as labeled samples during training. Such assigned labels, called pseudo-labels, are most commonly associated with the field of semi-supervised learning. In this work, we explore a broader interpretation of pseudo-labels within both self-supervised and unsupervised methods. After a thorough treatment of pseudo-labeling in these areas, we draw the connection between them and identify commonalities between fields, as well as new directions where advancements in one area would likely benefit others, such as curriculum learning and self-supervised regularization.
- 17 March 2026
- Published by AI Access Foundation in Journal of Artificial Intelligence Research
- Vol. 85
Top Cited PapersA Review of Pseudo-Labeling for Computer Vision
Deep neural models have achieved state-of-the-art performance on a wide range of problems in computer science, especially in computer vision. However, deep neural networks often require large datasets of labeled samples to generalize effectively. An important area of active research is semi-supervised learning, which attempts to instead utilize large quantities of (easily acquired) unlabeled samples. One family of methods in this space is pseudo-labeling, a class of algorithms that use model outputs to assign labels to unlabeled samples which are then used as labeled samples during training. Such assigned labels, called pseudo-labels, are most commonly associated with the field of semi-supervised learning. In this work, we explore a broader interpretation of pseudo-labels within both self-supervised and unsupervised methods. After a thorough treatment of pseudo-labeling in these areas, we draw the connection between them and identify commonalities between fields, as well as new directions where advancements in one area would likely benefit others, such as curriculum learning and self-supervised regularization.
- Open Access
- 13 August 2024
- Published by Association for Computing Machinery (ACM)
- p. 787-793
Impact of COVID-19 on Linguistic Expression of Depression in Online Communities
The COVID-19 pandemic has profoundly impacted mental health worldwide, potentially altering the linguistic expression of depression in online communities. This study investigates these changes by analyzing Reddit posts from depression-related and other mental health subreddits before and after the pandemic declaration. We employ a novel methodology combining advanced Natural Language Processing (NLP) techniques with statistical analysis. Fine-tuning the roBERTa model on the DepressionEmo dataset [24] achieves state-of-the-art performance in identifying depression-related emotions, with 80% precision, 82% recall, and 81% F1-score. Chi-square tests reveal significant post-pandemic increases in the expression of loneliness, sadness, suicide intent, and worthlessness (p < 0.001) in depression subreddits compared to other mental health forums. Anger and brain dysfunction exhibit divergent trends between the two groups. These findings underscore the pandemic's differential impact on depressive language, highlight the need for targeted online mental health support, and demonstrate the potential of interdisciplinary methodologies in analyzing mental health discourse at scale.
- 13 August 2024
- Published by Association for Computing Machinery (ACM)
- p. 787-793
Open AccessImpact of COVID-19 on Linguistic Expression of Depression in Online Communities
The COVID-19 pandemic has profoundly impacted mental health worldwide, potentially altering the linguistic expression of depression in online communities. This study investigates these changes by analyzing Reddit posts from depression-related and other mental health subreddits before and after the pandemic declaration. We employ a novel methodology combining advanced Natural Language Processing (NLP) techniques with statistical analysis. Fine-tuning the roBERTa model on the DepressionEmo dataset [24] achieves state-of-the-art performance in identifying depression-related emotions, with 80% precision, 82% recall, and 81% F1-score. Chi-square tests reveal significant post-pandemic increases in the expression of loneliness, sadness, suicide intent, and worthlessness (p < 0.001) in depression subreddits compared to other mental health forums. Anger and brain dysfunction exhibit divergent trends between the two groups. These findings underscore the pandemic's differential impact on depressive language, highlight the need for targeted online mental health support, and demonstrate the potential of interdisciplinary methodologies in analyzing mental health discourse at scale.