1 Overview

• Particling Night is an interactive media art project that invites viewers to remember others and reflect on the traces relationships leave behind. By translating the viewer’s name and voice into speech-to-text input, particle-based generative graphics, and the metaphor of stars and the universe, the work encourages a sensory reflection on human relationships.
• The project began as a 2022 undergraduate project for Creative Computing for Media Art at UNIST, later expanded into an exhibition at HCI Korea 2023, and was subsequently developed into a paper for IASDR based on exhibition observations and survey findings.
• Role: artwork concept development, interaction design, generative graphics and speech-recognition implementation, exhibition evaluation, and short paper writing

2 Context

Particling Night began from a concern that while social media expands our networks, it can also make our sense of relationship thinner and more superficial.

Today, we are more densely connected than ever through social media, yet people paradoxically report greater loneliness and isolation. Human relationships are increasingly remediated into discrete platform actions such as “follow” and “unfollow”. From this concern, the project asks what truly meaningful relationships are, and what kinds of traces those relationships leave on who we are now.

Particling Night is an interactive media artwork that invites viewers to reflect on the meaning of the relationships they form within the vast universe of social networks. Just as Yoon Dong-ju’s poem Counting the Stars at Night layers stars and the night sky with memory, longing, transience, and an ideal world, this project also uses the metaphor of the universe and stars to express relationships and their lingering traces. It proposes an experience through which viewers can reflect on their human relationships and cultivate self-resilience through artistic experience.

3 Approach

This project unfolded through a four-stage process, from early ideation and system implementation to an empirical user study conducted through exhibition.

3-1 Metaphorical Concept and Interaction Design

Rather than explaining relationships directly, the work was designed to let viewers speak names aloud and interpret the movement of stars for themselves.

“How have encounters and partings with others changed who I am now?”

From a distance, relationships can look like a repeated pattern of people gathering and dispersing. But from the perspective of the self, others seem to enter one’s life, pass by, and leave behind traces of varying depth. Instead of describing this feeling explicitly, the project was developed so that viewers would speak their own name and the names of others, watch the resulting stars move, and interpret those relationships on their own.

To support this reflective experience, the work drew on memory and ambiguity as key design strategies. Viewers recall the names of people they have encountered in both physical and digital life and speak them aloud, while the system transforms each name into an abstract, generative celestial form.

Relationships and interaction were metaphorized through gravity and celestial motion.

At the center of the screen, a star representing “me” appears, emphasizing that the experience unfolds from a deeply personal point of view.

Each time a viewer speaks a name, a celestial graphic appears on the right side of the screen. These bodies vary in form, position, and speed. Some brush past the central star, some linger in orbit for a long time, and some are eventually absorbed. This flow metaphorically captures the different ways people enter our lives, remain for a while, and drift away.

Within this zone of influence, stars exchange particles in each other’s colors, and when one is fully absorbed into the central star, the particle form of the central star changes. These transformations suggest that the traces left by relationships gradually reshape the self over time.

Through the metaphors of stars, orbit, and gravity, the work sought to communicate the contingency of encounters and the accumulation of influence without directly explaining the intensity or closeness of any relationship. This ambiguous visualization does not label relationships in fixed terms, but instead creates space for viewers to project their own deeply personal experiences onto the work.

3-2 Interactive System Implementation

Interactive system

• The work takes input through a keyboard and microphone. When viewers type their own name, a star representing themselves appears at the center of the screen. When they then speak the names of others into the microphone, each name generates a new particle-based celestial body.
• Voice amplitude was used to determine the size and style of graphic elements, and spoken names were recognized using the Kakao Speech-to-Text API. The recognized names were converted into hexadecimal color codes and used as the colors of the celestial bodies.
• The processed data was transmitted via UDP to Processing 4, where it functioned as parameters for real-time generative visualization.

Generative graphics implementation

• Existing open-source particle graphics were adapted and reworked for the project’s visual effects.
• The concept of gravity was incorporated into the visualization so that when surrounding stars entered a certain region around the central star, they accelerated toward it.
• Interactions between stars were translated into particle exchange, so that when stars approached within a certain distance, they exchanged colored particles or were fully absorbed.

3-3 Exhibition and User Study

The work was later exhibited at HCI Korea 2023 in Gangwon-do, South Korea, in February 2023. Visitors to the exhibition booth first received a brief explanation of the concept and interaction method, then experienced the work directly, and afterward participated in an anonymous survey. A total of 17 participants volunteered to respond, most of whom were Koreans in their twenties and thirties.

After the exhibition experience, participants completed a seven-item questionnaire using a five-point Likert scale. The questions addressed three broad areas:

• how participants perceived relationships in digital environments,
• whether the artwork actually prompted reflection on relationships,
• which elements among name vocalization, graphics, and animation served as triggers for reflection,
• and whether digital media cues or physical sensory cues played a stronger role in recalling people.

In addition to the survey, audience conversations and behavior were also observed on site. Some viewers interpreted the movement of generated stars in relation to their real-life relationships and discussed their thoughts with companions, while others hesitated to say names out loud even when someone came to mind. These observations complemented what could not be fully captured through survey scores alone.

4 Outcome

4-1 Generative Media Artwork: Particling Night

• Generative visualization that gives each experience uniqueness: When a viewer types their name, a uniquely colored central star is created. As the viewer speaks the names of people they know, particle-based celestial bodies with different forms and colors are generated on the right side of the screen based on the spoken names and voice amplitude.
• Interaction visualization that captures the traces of relationships: As viewers say names one by one, each name appears as a particle-based celestial body with its own color and form. These bodies pass by the central star, linger, and are sometimes absorbed. By watching their movement, viewers are invited to recall a particular relationship and quietly reconsider what that person has meant to them and what trace that relationship has left on their present self.
• Commemorating the experience: At the end of the interaction, the star bursts outward until it fills the screen and then freezes. A final screenshot is provided, recording the viewer’s name and the number of names they called as a caption, allowing the experience to be celebrated and kept as a memory.

4-2 Findings from the User Study

Taken together, the survey and observational findings suggest that the work functioned as a medium through which viewers could recall and reinterpret their relationships. Speaking names aloud, watching the generative graphics emerge, and observing the movement between stars did not operate as isolated features. Rather, they worked as one continuous experience that prompted reflection.

• The work as a whole was received as an experience that encouraged viewers to look back on their relationships.

  Participants generally responded that the artwork helped them think about their relationships again (median 4, mode 5). This suggests that the work functioned not merely as an audiovisual stimulus, but as a trigger for revisiting personal experience.

• Speaking names aloud functioned as a particularly strong reflective trigger.

  Among the different elements of the work, the act of vocalizing people’s names was perceived as especially effective in prompting reflection (median 4, mode 5). Rather than recalling relationships only internally, the embodied act of speaking became a starting point for activating memory.

• The graphics and movement created space for interpretation.

  The graphics representing people helped participants recall relationships (median 4, mode 5), and the movement between stars also worked as an interpretive cue (median 4, mode 4). Viewers watched stars pass by, linger, or become absorbed, and interpreted these scenes through the lens of their own experiences.

• Physical sensory cues played a slightly stronger role than digital media cues in recalling people.

  More participants responded that physical elements such as voice, scent, and gesture played a larger role in recalling someone than digital media elements such as profile images, photos, or emoticons (9 out of 17 participants). This suggests that what reactivates memory in relationships may not be digital information alone, but more embodied and sensory forms of remembrance.

Audience members responded to the generated scenes with comments such as, “Why does this person’s star pass by so quickly?” and “Why did this star stay so long?” Conversations naturally emerged in which viewers interpreted the movement of stars in relation to real-life relationships. This indicates that the work provided an interpretive openness through which viewers could layer their own memories and meanings onto the experience.

5 Reflection

• A gap between artistic intention and audience entry into the work:

  Although the system itself was simple to operate, many viewers in the exhibition setting did not fully read the instructions or wanted to see the result as quickly as possible. This made clear that in interactive systems, it is just as important to consider how audiences enter into the work as what the work ultimately shows. Meaning emerges only when viewers can understand and follow the intended experiential flow.

• Privacy as a condition for immersion:

  Because the exhibition took place in a public setting, some viewers were hesitant to say other people’s names out loud. This suggested that creating emotionally immersive experiences of reflection requires careful attention not only to visual design, but also to the degree of privacy afforded by the exhibition environment, or to the possibility of extending the experience into a more private online setting.

• A caution about how technology shapes interpretation:

  Speech recognition served as an intriguing entry point into the artwork, but for some viewers it also made the artwork seem like a system that was analyzing their emotions or relationships rather than presenting an ambiguous metaphor. This revealed an important design consideration for AI-based media art: when advanced technologies are incorporated into artistic experiences, the perceived analytical power of AI can flatten ambiguity and reduce space for reflection unless the interaction is designed with great care.