Creative Computing Graduate with interest in Procedural Generation and Emergent Behaviour, pursuing applications of AI and Algorithms for world building, immersive experiences coupled with AR and XR technologies. Passionate about exploring areas of music, fashion, and design particularly experimental, with ability to explore the interdisciplinary space of computer science and arts to offer novel perspectives.
PROJECTS
Matvey Iliyashov - Portfolio
(2024) PuppetParlour
Tools: OpenFrameworks, MediaPipe, Blender, OSC.
Phygital (physical and digital) experiences allow digital and physical systems to intertwine to create immersion through direct physical interaction with the digital world. Puppet Parlour examines this link by emulating the analog experience of controlling a hand puppet using media pipe, while simultaneously exploring the relationship between puppet and puppeteer, by asking “How can the analog experience of controlling a hand puppet, be mapped and subverted in a digital world?”.
(2025) Deep Dream Dungeon
Tools: Three.JS, MediaPipe, Blender, Stable Diffusion, Deep Dream, MusicGen, Inpainting.
Deep Dream Dungeon is a simulated multidimensional walking simulator where the player traverses various pre-generated deep dreamified environments using a portal gun in addition to using their face to control their destination. The project's goal was to create an explorable environment using stable diffusion whilst using face-tracking to let the player take control of the surrounding environment.
(2025) Final Funhouse
Tools: Unity.
Developed a three-dimensional procedural dungeon algorithm based on binary space partitioning. The Unity package is a tool for my artistic practice, generating 2D-3D layouts with parameterised room and dungeon sizes, positions, and textures through room blueprints. The project aims to develop a navigable 3D generator alongside a modular decoration system.
Recipient of best project Award across Creative Computing.
Developed experimental THREE.js frontend for LLM with progressively degenerative response through procedural prompt engineering. Backend using Node.js and Express.js with python audio and TTS processing, text-phoneme and audio-timestamps. Communication between backend and frontend achieved via socket.io.