
Dream Pipeline

Dec - 2025
The screen continuously outputs imagery generated from live speech. Spoken input is captured by a microphone and streamed through whisper.cpp for transcription. A Python script segments the text over time and uses a hand-built dream dictionary to assign thematic and emotional tags, expanding them into prompts sent to TouchDesigner to drive a real-time diffusion system. The work does not aim for semantic fidelity; instead, speech rhythm, pauses, and volume steer how the visual field evolves.
A camera simultaneously tracks movement in the space, leaving brief afterimages and traces in the generated output.
These traces alter the diffusion behavior by shaping where and how noise is injected, allowing bodily motion to enter as a perturbation rather than a depicted subject. Fogging and glitch layers are added at the output stage, with glitch intensity mapped to volume—louder voices produce stronger ruptures, quieter speech yields subtler distortion. The result is a continuously shifting perceptual field instead of a stable representation.
![]() | ![]() | ![]() |
|---|---|---|
![]() | ![]() | ![]() |
Type: Generative video system
Authors: Liu Qianhua (Leo) Date: Dec 2025
Dimensions: Variable Size
Materials: Microphone; Camera;
whisper.cpp; Python; TouchDesigner;
StreamDiffusion; Display.
Image Forming Apparatus
April 2025
Type: Interactive installation
Authors: Liu Qianhua (Leo); Jubelle Anne;Shelby Murchie
Dimensions: Variable Size Materials:
Bare Conductive; Max/MSP; Arduino;
Portable DVD player; Steel grid; VHS
tape; Cables; Projector; Speakers
"A constellation of dismantled media components—VHS tapes, cables, and circuit fragments—hangs beneath a welded steel mesh, hand-stitched into a “re-sutured memory fabric.” A projected global e-waste heatmap falls across the installation, situating these physical remnants within a landscape of discarded data. Viewers are invited to touch exposed metallic parts from decommissioned devices; linked through conductive materials, these fragments function as capacitive touchpoints that trigger sound samples. Children’s laughter, distant city streets, and animal calls serve as mnemonic cues—traces of the materials’ “former lives” that evoke the everyday environments in which such media once circulated.
At the heart of the work is the tension between “cold media” and “warm memory.” Through physical contact, viewers “awaken” and “excavate” buried emotional resonance from inert matter. The sonic experience is non-linear: an interaction driven generative soundscape in which multiple touch points allow fragments to be recombined into a shifting sonic montage.
Each encounter produces a unique polyphonic layering—like a chance symphony assembled from the ruins of memory. Through this “symphony,” e-waste is reimagined as a spatial, multi-dimensional field of feeling.






Mood Diary
Jan 2025
Type: AI-generated visual
installation
Author: Liu Qianhua (Leo); Luo Wen (Winnie)
Date: Jan 2025.
Dimensions: variable size
Materials: Autolume; Max/MSP; 100 sky photographs (Toronto); monitor; picture wall; AI-
generated video; readymade elements
Mood Diary investigates the
artist's agency in the age of artificial intelligence. Instead of using pre-trained AI as a passive tool, the work focuses on how the selection of training data serves as a primary creative intervention.

A grid of 100 photographs of the Toronto sky—capturing personal, emotional moments—serves as the foundational dataset. This archive is channeled through a custom-trained AI model to generate ambiguous topographies that oscillate between memory and abstraction. By blurring these sky-scapes, the work moves from individual nostalgia to collective memory, allowing viewers to project their own experiences onto the visuals.
To link this digital stream with the physical body, the output’s brightness and color are modulated by pre-recorded heartbeat data (integrated via Max/MSP).The core of the work is this negotiated uncertainty: exploring how AI can generate unexpected visuals while remaining anchored to the artist’s intent. It probes the liminal space between human authorshipand machine autonomy.


_JPG.jpg)
Spiritual world
3 January - 20 June 2019
Silicone · Video · Sound ·
LEDs · Mechanical Device ·
Readymade
"Spiritual World" simulates the scene of the operating room. Through the interaction of various factors, the formed spatial composition creates a plot that is contrary to our collective memory, gives the audience a psychological experience that is contrary to daily experience, and conveys A sense of absurdity, a sense of astonishment. Thus forming a dislocated overall framework to promote the extension of the concept.
.png)

Washing Machine
3 October - 30 December 2018
Washing machine · Video ·
Table and chairs
The author intends to dissolve the affiliation between the body and the washing machine in daily life, and construct a new context: to establish a new way of viewing and interpreting the washing machine through relationship reorganization.
It makes the waste washing machine avoid the destination of the garbage dump, and turns from garbage into art. The spatial composition relationship formed by it gives people a sense of religious ceremony, thus making the waste washing machine sacred and solemn.





Resin · video · sound
Readymade
MEMORIES
3 January - 20 June 2016
This intimacy makes the objects an extension of my grandmother's body, and at the same time combines the images of my grandmother holding these objects and recalling the past to reinforce the idea she wants to express.
In the process of creation,
I collected representative items of my grandma at various stages. Behind each ready-made product, there is a period of my grandma's life experience. These objects carry the grandma's personal memories and personal feelings, and the objects establish a specific memory and irreplaceable secret relationship with the grandma's body.

3 July - 20 October 2018
Mechanical device
Book
Silicome
Dream
A contrasting effect is presented through the combination of figurative and realistic sculptural language and ready-made books.
The ignorant little baby slept peacefully, but he didn't know that his future was clearly arranged.
The author wants to present it through a combination of "babies with simple colors" and "ready-made books with colorful colors". The baby and the "huge book" create a visual contrast effect from color and volume.




LEDs · video · sound ·Steel plate
mechanical device ·
Dislocation
3 January - 20 June 2020
2020 515 Art Creation Cave Art Museum
The work "Dislocation" is not just a stimulus that is satisfied with the audiovisual, it urges the audience to combine time, memory and the comprehensive experience of the scene, to reflect and question.

Biiiiird
Generative art
I attempted self-learning and exploration of Grasshopper for art creation, and sensed that generative art differs from traditional sculpture creation in terms of real-time feedback, ease of iteration, flexibility, and scalability. Grasshopper excels at manipulating curves and surfaces, facilitating a quicker and more potent art creation process by handling complex and repetitive shapes effortlessly within a short timeframe. Through the exploration of generative art, I've discovered that it contributes to broadening the dimensions and presentation possibilities of the creations.




















































_JPG.jpg)
