Voicescapes (2025)
Interactive Art
System Design, AI Pipeline Development, Interaction & Hardware Fabrication
➊ Speech Recognition & Text Transcription: Whisper (OpenAI) / Python
➋ English Prompt Generation: OpenAI GPT API / Python
➌ Image Generation: StreamDiffusionTD
➍ Real-Time Interface & Video System: TouchDesigner
➎ Hardware Fabrication: Arduino / 3D Printing / Cinema 4D
Interactive Art
System Design, AI Pipeline Development, Interaction & Hardware Fabrication
➊ Speech Recognition & Text Transcription: Whisper (OpenAI) / Python
➋ English Prompt Generation: OpenAI GPT API / Python
➌ Image Generation: StreamDiffusionTD
➍ Real-Time Interface & Video System: TouchDesigner
➎ Hardware Fabrication: Arduino / 3D Printing / Cinema 4D
A voice becomes a sentence, a sentence is translated into an image, and the image flows across time.
In the exhibition space, an automated system was built to extract text from the audience’s spoken voice through a microphone and use AI models to generate images and videos in real time. Ultimately, this project realizes a real-time interactive video system in which the viewer’s voice is transformed sequentially into text, images, and moving scenes—forming a continuous flow from speech to time-based visual experience.
In the exhibition space, an automated system was built to extract text from the audience’s spoken voice through a microphone and use AI models to generate images and videos in real time. Ultimately, this project realizes a real-time interactive video system in which the viewer’s voice is transformed sequentially into text, images, and moving scenes—forming a continuous flow from speech to time-based visual experience.