Welcome to the South Caucasus Conference on Artificial Intelligence!
We are delighted to announce that the South Caucasus Conference on Artificial Intelligence (SCCAI) will be held in Tbilisi, Georgia from September 16 to September 18 2025. The event is organized by the Muskhelishvili Institute of Computational Mathematics (MICM) of the Georgian Technical University (GTU), the German Research Center for Artificial Intelligence (DFKI) and the National Institute for Research in Digital Science and Technology in France (INRIA).
The conference aims to create an innovative, cross-disciplinary platform where academics, industry experts, and government officials unite to explore new frontiers in theory, methodology, systems, and applications of Artificial Intelligence (AI). This year, the focus is on the latest developments in Computer Vision and Natural Language Processing for Low-Resourced Languages.
SCCAI 2025 seeks to establish a collaborative research platform for the South Caucasus nations – Armenia, Azerbaijan, and Georgia, promoting their integration into the global AI community.
The conference is organized and fully financed by the European Union HORIZON EUROPE Project GAIN https://www.gain-twinning.eu/
No conference fee is needed.
SCCAI2025 invites researchers to submit abstracts for talks as well as full papers to one of two thematic conference tracks and Doctoral Symposium:
NATURAL LANGUAGE PROCESSING FOR LOW-RESOURCED LANGUAGES
September 18
Track chairs: Jan Alexandersson & Philipp Müller – DFKI, Beso Mikaberidze – MICM
The South Caucasus region is home to a variety of small languages with relatively few speakers. This poses a significant challenge to NLP researchers and engineers because large language corpora – that are needed to train high-performing NLP systems – are scarce. In order to tackle this challenge, Track 1 focuses on NLP for low-resourced languages. Topics of interest include, but are not limited to:
- New corpora for low-resourced languages
- Tokenization and POS tagging in low-resourced languages
- LLMs for low-resourced languages
- Machine translation and cross-lingual transfer learning for low-resourced languages
- Speech recognition and synthesis for low-resourced languages
- Data augmentation and synthetic data generation for low-resourced languages.
COMPUTER VISION FOR PERCEPTION, INTERACTION & INTELLIGENT SYSTEMS
September 18
Track chairs: François Brémond & Michal Balazia – INRIA, Teimuraz Saghinadze – MICM
Vision is the primary sense used by human individuals to interact with their environment. Consequently, the field of computer vision offers a variety of promising applications: From autonomous cars to new ways of person identification and intelligent support systems for healthcare professionals. Thus, Track 2 focuses on various fields of computer vision that include, but are not limited to:
- Analysis of human interaction and emotion (e.g., backchannel detection, arousal classification, gaze estimation, intent prediction)
- Person identification (e.g., face recognition, gait recognition, gesture-based identification)
- Computer vision in highly dynamic environments (e.g., autonomous driving, human-machine collaboration in manufacturing and logistics, etc.)
- Computer vision for healthcare and assistive technologies (e.g., patient monitoring, activity recognition for rehabilitation, sign language recognition
- Multimodal augmentation of computer-vision approaches (e.g., audio, physiological data, etc.)
- Explainable and ethical AI in computer vision (e.g., bias mitigation, interpretability of deep vision models, fairness in person identification).
Doctoral Symposium
September 16
Track chairs: Tanay Agrawal – INRIA & Ioseb Kachiashvili – MICM
The Doctoral Symposium will give PhD students working in any AI-related area an opportunity to share their ongoing research work and ideas with renowned and experienced researchers in the field and get suggestions and feedback from them.
The students are encouraged to present their research at any step of the scientific process (e.g., conceptualization phase, programming phase, training phase, evaluation phase, etc.)
PhD students should send a short paper (up to 5 pages including 1 page for references) summarizing their current research. A program committee will review the submissions. The accepted papers will be presented by the students to a panel of senior experts (chosen by the Doctoral Symposium track chairs) who will provide constructive feedback and suggestions.
