Andreas Panayiotou

I'm

Short Bio

Andreas Panayiotou Profile Picture

Hey. I am Andreas and I was born in Limassol, a beautiful city in Cyprus. Currently, I am a PhD student at University of Cyprus and a research associate at CYENS - Center of Excellence. In 2022, I received my M.Sc in Computer Science from University of Cyprus, while also I hold a B.Sc in Computer Science from the same university. My interests include, but not limited to, Crowd Simulation, Machine Learning, Virtual Environments, Computer Animation and Computer Games Technologies.

Resume

Full CV - Available Here

Education

Doctor of Philosophy (PhD) Computer Science

Sep 2022 - Present

University of Cyprus (UCY), Nicosia, Cyprus University of Cyprus Logo

Details

Topic: Populated Virtual Environments

Lab: UCY Graphics Lab

Master of Science (M.Sc) Computer Science

Jan 2021 - Jun 2022

University of Cyprus (UCY), Nicosia, Cyprus University of Cyprus Logo

Details

Courses: Modern Technical Analysis of Large Scale Graphs, Distributed Systems, Visual Computing, Video Game Software Development, Cognitive Programming, Advanced Software Engineering, Electronic Health

Thesis: Crowd simulation by Deep Reinforcement Learning - Available Here

Bachelor of Science (B.Sc) Computer Science

Jan 2017 - Jan 2021

University of Cyprus (UCY), Nicosia, Cyprus University of Cyprus Logo

Details

Elective Courses: Synthesis of Parallel Algorithms, Computational Intelligent Systems, Logic in Computer Science, Advanced Networks, Network and Information Security, Internet Technologies

Thesis: Shift scheduling software and mobile application for Ambulance Service Cyprus

Professional Experience

Visiting PhD Student

Sep 2024 - Present

UC Riverside | Motion Planning Lab, Riverside, California alt="UCR

Details
  • Crowd Simulation
  • Generative AI
  • Graph Machine Learning
  • Computer Graphics

Research Associate

Sep 2021 - Present

CYENS - Centre of Excellence | V-EUPNEA MRG, Nicosia, Cyprus

Details
  • Computer Graphics
  • Crowd Simulation
  • Machine Learning
  • Game Technologies

Research Intern

Jun 2021 - Jul 2021

CYENS - Centre of Excellence | V-EUPNEA MRG, Nicosia, Cyprus

Details
  • Unity | ML-Agents Toolkit
  • Training of agents using Imitation and Reinforcement Learning
  • Character Animation

Information Technology Trainee

Jun 2018 - Jul 2018

Bernhard Schulte Shipmanagement, Limassol, Cyprus Bernhard Schulte Shipmanagement Logo

Details
  • Support company core network
  • Provide technical support to employees
  • Communicate with international company partners to learn the use of maritime softwares
  • Upgrade and repair of computer hardware

Selected Publications

2022
CCP: Configurable Crowd Profiles

Andreas Panayiotou, Theodoros Kyriakou, Marilena Lemonari, Yiorgos Chrysanthou, and Panayiotis Charalambous

In Special Interest Group on Computer Graphics and Interactive Techniques Conference Proceedings (SIGGRAPH ’22 Conference Proceedings)

CCP: Configurable Crowd Profiles
Diversity among agents' behaviors and heterogeneity in virtual crowds in general, is an important aspect of crowd simulation as it is crucial to the perceived realism and plausibility of the resulting simulations. Heterogeneous crowds constitute the pillar in creating numerous real-life scenarios such as museum exhibitions, which require variety in agent behaviors, from basic collision avoidance to more complex interactions both among agents and with environmental features. Most of the existing systems optimize for specific behaviors such as goal seeking, and neglect to take into account other behaviors and how these interact together to form diverse agent profiles. In this paper, we present a RL-based framework for learning multiple agent behaviors concurrently. We optimize the agent by varying the importance of the selected behaviors (goal seeking, collision avoidance, interaction with environment, and grouping) while training; essentially we have a reward function that changes dynamically during training. The importance of each separate sub-behavior is added as input to the policy, resulting in the development of a single model capable of capturing as well as enabling dynamic run-time manipulation of agent profiles; thus allowing configurable profiles. Through a series of experiments, we verify that our system provides users with the ability to design virtual scenes; control and mix agent behaviors thus creating personality profiles, and assign different profiles to groups of agents. Moreover, we demonstrate that interestingly the proposed model generalizes to situations not seen in the training data such as a) crowds with higher density, b) behavior weights that are outside the training intervals and c) to scenes with more intricate environment layouts.
SIGGRAPH 2022 Logo

Projects

SHARESPACE.eu
sharespace.eu
The vision of SHARESPACE is the creation of future Social Hybrid Spaces (SHS) shared by humans and avatars engaged in embodied collaborative tasks, where social sensorimotor primitives are transparently captured through mobile connected innovative sensors, and then reconstructed using novel extended reality (XR) technology. Our ambition is to create a hybrid, multimodal-multisensory integrated platform which adapts to individual users and enables them to interact in an embodied shared space by learning, identifying, and reconstructing the core sensorimotor primitives of social interactions. Through our cognitive architecture, virtual avatars will facilitate joint-action, distance-based learning, and social cohesion in the hybrid societies of the future, by reconstructing core social sensorimotor primitives into virtual ones. Our long-term vision is to expand the natural repertoire of sensorimotor interactions as we currently know it, with XR technology enriching key social features of sensorimotor communication tailored to the participants’ personal characteristics, with future deployment in sectors such as health, learning, entertainment, sport and at the workplace. In short, the project aims to bring XR to a radically, ethically sound, new level of presence and sociality by furthering the possibilities of interaction in a virtual space using avatars.
ReInHerit.eu
ReInHerit - Redefining the Future of Cultural Heritage
ReInHerit is an Horizon2020 project that aspires to disrupt the current status quo of communication, collaboration and innovation exchange between museums and cultural heritage sites, in a sense that it will connect cultural heritage collections and sites, and present Europe’s tangible and intangible heritage to citizens and tourists in their wider historical and geographical contexts. The ReInHerit project is proposing a very innovative model of sustainable heritage management, through which a dynamic network will be born; this network comprised of cultural heritage professionals, innovation and cultural heritage solution tech experts, researchers, national museums, regional and local museums, and representative managers of Heritage Label sites. ReInHerit is also supported externally, by relevant municipalities, regional and local museums, which will also enhance the communication and cooperation of the cultural heritage sector across the EU and beyond.

Contact

Location:

Dimarchias Square 23, Nicosia, 1016, Cyprus