Stefanos Gkikas

I am a Ph.D. candidate focusing on Affective Computing and Emotion AI. My doctoral research specifically focuses on automatic pain assessment using multimodal data sources. Currently, I am collaborating with the Biomedical Informatics & eHealth Laboratory of the Hellenic Mediterranean University and with the Computational BioMedicine Laboratory (CBML), part of the Institute of Computer Science at FORTH.

I am interested in emotion recognition and human behavior analysis, employing advanced deep-learning methods for video data and biosignals.

Email  /  CV  /  Scholar  /  GitHub

Research

The proposed multimodal framework utilizes facial videos and fNIRS and presents a modality-agnostic approach, alleviating the need for domain-specific models. Employing a dual ViT configuration and adopting waveform representations for the fNIRS, as well as for the extracted embeddings from the two modalities, demonstrate the efficacy of the proposed method.

This study presents synthetic thermal videos generated by Generative Adversarial Networks integrated into the pain recognition pipeline and evaluates their efficacy. A framework consisting of a Vision-MLP and a Transformer-based module is utilized, employing RGB and synthetic thermal videos in unimodal and multimodal settings.

The proposed framework comprises four pivotal modules: the Spatial Module, responsible for extracting embeddings from videos; the Heart Rate Encoder, tasked with mapping heart rate signals into a higher dimensional space; the AugmNet, designed to create learning-based augmentations in the latent space; and the Temporal Module, which utilizes the extracted video and heart rate embeddings for the final assessment.

Presenting a full transformer-based framework consisting of a Transformer in Transformer (TNT) model and a 1D Transformer leveraging cross-attention and self-attention blocks for video analysis.

Introduction of a novel multi-task neural network for automatic pain assessment utilizing the age and gender information of each individual alongside the pain estimation.

This review aims to identify original studies based on deep learning approaches and discuss the models, the methods, and the types of data employed in establishing the foundation of a deep learning-based automatic pain assessment system.

Extracting hand-crafted features from electrocardiography signals with machine learning algorithms and exploring the correlation of gender and age with the manifestation of pain.