Art Fund Project
Tango Wish AI
Interactive AI part
The second part invited the audience to create prompts and transform themselves into different characters using generative AI. Placing these two approaches side by side allowed me to explore how people respond to human-made versus AI-generated visuals, while also questioning which creative process brings me more joy as an artist.
Technical summary
I found working with body tracking and LFO systems the most satisfying because there was a more direct relationship between my creative decisions and the visual outcome. StreamDiffusion, in contrast, involved extensive debugging with Claude and often felt closer to software engineering than visual design. This made me realise that what I value most is not necessarily whether a tool uses AI, but how directly I can shape and respond to the outcome. I also aimed to use Control Net, the network architecture that allows more precise control over AI models but it unfortunately failed despite extensive debugging. My next steps is getting more experience working with Microsoft Kinnect-camera system that sees and tracks people in complete darkness
Tools
Nvidia Flow, Nvidia Background removal, Particle GPU system, Body Tracking, LFO system, Tops Filters, Geometry Instancing with a bit of python, 360 projection mapping, webcam input, Stream Diffusion and Node based system built from scratch and Motion Capture
Overview
This immersive dance performance combined movement, interactive visuals and AI within a 360° animated space. Developed over three months with support from the Immersive Arts Explore grant, I took the project from initial concept and experimentation through to production and exhibition at Not For Sale Gallery.
Human Art Part
The first part explored visuals and interactive systems I created myself. The dancer moved through responsive particles, fire, lightning, digital plants and human-shaped data, while gestures could paint directly onto the walls and control elements of the environment. I wanted the body to become an interface and the technology to feel playful, physical and alive.
Collaborators
-immersive Arts Fund
-George Pinnegar (FreshLov3) DJ and Gallery Curator
-Not for Sale Gallery
-Valentina Gonzalez-contemporary style dancer, performance artist.
-Amz-DJ
-Alexa Morena-sculpture
Donut Jam
Concept: Stage visuals and audioreactive animation for a jazz jam.
I designed colourful liquid visuals and audioreactive music notes in TouchDesigner, then projection mapped them across the stage and surrounding space. I wanted the experience to feel theatrical, so I highlighted the curtains with animated red textures and transformed the canvases above the audience into living paintings. The canvases became one of the most successful elements of the installation and particularly resonated with the organisers.
Technical: I built the liquid textures in Touchdesigner using GSLS textures and a bit of python. Audioreactive music notes were built using partincle GPU system. The visuals were output to 2 projectors and projection mapped on stage
Collaborator: Staffordshire St Gallery Donut Jam
Immersive Meditation
Technical: I created the audioreactive stars and GSLS texture from scratch and then projection mapped evrything using Touchdesigner and a single projector.
Collaborator: Lihui Liang
Concept: Immersive Audio Visual meditation for Lihui Liang’s Tiktok lifestream. The experience was designed to feel enigmatic and evocative for those participating in soundbath while the preparation didn’t take much time compared to my other projects, 2 hours to create visuals and 30 minutes setting things up.I projection mapped audioreactive stars behind and the water swirl visual on Lihui's white dress. The water motion was mirroring her circular movements while the stars were lightting up with every bang sound. This cosmic theme highly resonated with Lihui. For the second part of the experience I only kept the water swirl visual highlighting Lihui's white dress and presence even further.
Exploding Cinema
Concept: Animated canvases
This work was part of Exploding Cinema Film Festival. Installation Concept: psychodelic animated canvases. I created infinitely looped fractals and other black and white abstract visuals including audioreactive cloud in Touchdesigner. Then I projection mapped them on a mirror and canvas creating living paintings. I aimed to keep it minimalistic in palette and shapes to hopefully achieve enigmatic feel and to bring the focus to the subject of the canvas and mirror.
Prior to the exhibition, during the visual research stage I tested the audioreactive particle GPU cloud with input audio, creating a retro edit to juxtaposition retro and futuristic. This video was a blueprint for the earie atmosphere I was aiming to create with this installation. However, I think audioreactivity worked better with the input audio.
Collaborators: Exploding Cinema Film Festival, Set Social Pecham
Tower Collective
Concept: Live Poem animation for Tower Collective's Jewellery Exhibition For The Seeds To Bloom intended to contribute to St Anne's Church overall nostalgic atmosphere and medieval feel during the run time of the exhibition and performance and to display Stella's and Kong's poem alongside it. I also created the retro looking narrative social media edit.
The main challange with this project was the technical set up as the surface I worked with was bricks and there was no obvious place to put my projector and I needed to avoid visitors blocking the light.
Result: The poem animation was playing on white fabric alongside Stella's and Kong's video above jewellery installation. The poem looked washed out and not immediately readable as intended.
Collaborators: Maria Stella Lydaki -jeweller and video creator. Kong Qianyang-video creator and poem writer. Johanna Glaza-performance artist feauring in the video.
As the exhibition took place in Church, the setting informed the handwritten ink bleed style of the text animation, classic looking font choice and the black and white pallette.
Technical: I see the work I created as immersive motion graphics piece, as it's a poem animation and Touchdesigner was used only for projection mapping. I imported Illustrator templates inside After Effects and animated the text with a custom built ink bleed effect. Then I created 2 visual style version of the poem animation: one with texture behind the text for social media and another with a plain white background for projection as the fabric was the texture.
Software: Illustrator, After Effects, Touchdesigner