Sihwa Park

About

Selected Works
Diffusion TV
YouTube Mirror
Uncertain Facing
GeoD
ARLooper
Ballet Mécanique
Brand Logo Sonification
TopicBubbles
Loading...
BeHAVE
InstaSynth
Structured Improvisation X
FormSound

Archive
SoniPi
Letter Frequency Visualization
Generative Audiovisual Study
Ring of Quartets
Hans Zimmer’s OST vs Film
︎︎︎Past Works (2008-2015)

Recent News

- Dec 11, 2024: Diffusion TV has been accepted for the ISEA 2025 Exhibition!
- Aug 30, 2024: I performed in Exit Points#52 at Arraymusic, Toronto.
- Apr 7, 2024: YouTube Mirror was accepted for CVPR AI Art Gallery (Online)!
- Oct 24, 2023: I was appointed a Connected Minds Research Enhanced Hire position with a $100,000 startup fund.
- Sep 1, 2023: YouTube Mirror was presented at AIMC 2023, Brighton, UK.
- Jun 12, 2023: YouTube Mirror was accepted for AIMC 2023 as both a paper and an installation!
- May 31, 2023: YouTube Mirror was exhibited at NIME 2023, Mexico City, Mexico.
©2024 Sihwa Park

Sihwa Park


(WIP) Diffusion TV

Date: Work-in-Progress (Since May 2024)
Categories: AI Art, Audiovisual Installation, Generative

Diffusion TV is an experimental and interactive interface that invites audiences to explore the hidden processes of AI through a nostalgic CRT TV interface, allowing them to engage directly with diffusion models, the technology behind today’s most sophisticated AI-generated images and sounds.

Inspired by Nam June Paik’s Egg Grows (1984–1989), which depicts the illusion of an egg growing through a series of progressively larger CRT TVs, Diffusion TV uses the old CRT TV to reveal the hidden process of AI denoising, transforming the complexity of AI diffusion models into a tangible, interactive experience. The work allows viewers to directly engage with the transformative process of turning chaotic data into recognizable images and sounds by adjusting the TV’s antenna and knobs, fostering a deeper connection between human interaction and technological transformation.

By turning the knob, viewers can switch between three channels: Past, Present, and Future. The Past channel focuses on extinct animals to reflect on what has already disappeared, evoking a sense of loss and remembrance. The Present highlights the fragility of endangered species, offering a moment of reflection on the current environmental crisis. The Future channel displays speculative animals imagined with AI, encouraging contemplation about where humanity and technology are headed. Each channel presents audiovisual content generated by AI models such as Stable Diffusion XL and Stable Audio Open, dynamically adapting to the viewer’s adjustments.

As a work-in-progress, Diffusion TV currently presents a single Present channel, showing endangered animals identified by the World Wildlife Fund, an international conservation organization dedicated to protecting the world’s most vulnerable species and habitats. By adjusting the antenna, the audience gradually brings the endangered animals into clearer focus, metaphorically preserving their fading existence. Once the clearest image and sound of an animal are revealed, the process resets after a few seconds, returning to a noisy and unclear depiction of another animal, keeping the experience dynamic and ever-evolving. Like television programming, Diffusion TV continuously renews its content by regenerating the images and sounds of the animals at regular intervals.

Diffusion TV explores how technology can shape our understanding of the natural world. Using the CRT TV, once an emblem of cutting-edge innovation but now a metaphor for obsolescence, the work mirrors the fragility of endangered species and asks what is preserved and what is lost in the pursuit of technological advancement. Through a more experiential, interactive understanding of AI’s inner workings, Diffusion TV encourages audiences to contemplate the interconnected relationships between humanity, technology, and nature.