computer vision ART installation

Face to Interface 

An interactive art installation powered by Artificial Intelligence

My role
I worked as Lead User Interface Designer and Creative Director

Final Master’s Degree brief: make interactive, machine-powered art with a lean, super-tuned team, to be shown as a pilot project at a sensational new exhibition at the Vancouver Art Gallery.

 

WHY AI?

Artificial intelligence is rapidly becoming an ever-present technology in the creative, academic, and professional environment, yet human intelligence must more critically assess the potential impact of an AI driven world. The Imitation Game: Visual Culture in the Age of Artificial Intelligence was a pertinent 2022 Vancouver Art Gallery exhibition that allowed visitors to experience deep learning production processes first hand, and encouraged informed discourse around the use - and misuse - of such in all areas of Visual Culture.

↑ Team 4149+i, generated with virtual avatar app Pinscreen

Team 4149+i, generated with AI face synthesis technology app Pinscreen

 
 
 

HELLO WORLD

From initial conversations with our client we had established our target audience, a broad spectrum of four ‘types’:

  • Novice attendees with interest in, but little to no prior knowledge of the subject matter

  • Organization members and culture enthusiasts 

  • Visitors with higher education, perhaps in areas such as art, computer science or related subjects

  • Machine intelligence researchers, professionals and industry peers

 
↑ The Eye: face recognition

↑ ‘The Eye’ exhibit interface concept

 
 

REVEALING EMOTIONS

The prototype my team created allowed visitors to interact with The Eye - AI software that reads facial expressions, then gauges related human emotional states. Once a face was read, the interface responded by mirroring the perceived emotion with an output of images pulled from its own extensive training dataset.

 

↑ Live exhibit footage

 
↑ onboarding interaction flow, from first face scan to emotion recognition

↑ onboarding interaction flow, from first face scan to emotion recognition

 

INCEPTION

To be able to create an appropriate interface for such an innovative show, I had to first discover what the future of cultural spaces could look like in a post-pandemic world. I found inspiration in viewing rooms, art fairs and virtual art conferences, social media casts and digital art marketplaces such as Mozilla Hubs, Somnium Space, the VOMA, Recto VRso, Laval Virtual, Frieze, ArtGateVR, Hauser & Wirth HWVR, CannesXR, Helsinki Fashion Week, CADAF Online, the MOR, the DiMoDa, Epoch Gallery, as well as virtual worlds powered by the blockchain such as SuperRare, Decentraland, RareAF, Async Art, Cryptovoxels 🧬 and many, MANY more.

 
 
↑ Decentraland

↑ Decentraland

↑ Cryptovoxels / SuperRare

↑ Cryptovoxels / SuperRare

↑ Helsinki Fashion Week

↑ Helsinki Fashion Week

↑ Cryptovoxels

↑ Cryptovoxels

↑ Mozilla Hubs / the Circle VR party

↑ Mozilla Hubs / the Circle VR party

↑ Epoch Gallery

↑ Epoch Gallery

 
↑ CADAF Online

↑ CADAF Online

↑ Recto VRso

↑ Recto VRso

↑ VOMA

↑ VOMA

 

STRICT MACHINES

In order to present viewers with a satisfying reading of the complex layering of their emotional states we looked into employing cutting-edge sophisticated face tracking and emotion analysis technology. 

We learned about affective computing – the way technical sensing systems and devices are learning to observe, read and understand human feelings. To teach a machine to understand emotional states, it can simply observe a human face.

Facial recognition technology can be used to classify expressions – a smile, frown or scowl – and to infer the emotional state of a person, such as ‘happy’, ‘sad’ or ‘angry’. It can be used from a distance, and can be performed without the knowledge of the individual. 

 
↑ emotion detection via facial feature data points

↑ emotion detection via facial feature data points

 
 
 

LOOK & FEEL

The message and visual language of this particular User Interface was to be keen machine intelligence, in observation mode. I opted for strict monochrome HUD treatments, and a disorienting blur to the ambient shadow a human body generates, that would entice users to step closer to investigate.

  • Futuristic and unfamiliar

  • Geometric and precise

  • Mechanical and automated

  • Mysterious and somewhat unsettling

 

DESIGN PRINCIPLES

We needed to plan a sophisticated experience for the modern day museum visitor: portray and balance AI lifestyle benefits such as convenience, safety, fun and familiarity; with it’s innate risks, such as data exposure, surveillance and algorithmic brainwashing.

We intrinsically needed to ensure people would feel safe and comfortable once engaged in the installation, yet work on retaining an element of surprise.

The agreed upon design principles, established in close collaboration with our clients, were:

  • Visually engaging

  • Rich and diverse

  • Intellectually rigorous

  • Sophisticated and satisfying

 
 
↑ first test view on a local PC
 
mars-63034_1280.jpg
 

PROGRAMMED FEELINGS

Each visitor was guaranteed a playful experience, mapped to their individual emotions. Yet, whilst believing themselves in control, users were also freely relinquishing biometric data, parsed from the very first moment of engagement.

↑ there is no spoon
 
 

DEUS EX MACHINA

The Cyber Eye signified the beginning of the journey through this experience, and sat atop the stylised interfaces of four perpendicular LFDs, on a rigid 4-sided plinth, ever watchful for human presence. It served as a sentinel, biding its time, idly sending deep environmental space scans out to the foremost reaches of its vision, searching for human presence.

The Eye utilises the visual language of sacred geometry and divine symbolism - a Machine God, casting its vigilant eye over the world. The history of semantics would present this magical sigil as a sign of protection, good health and providence. But for whom - human, or machine?

 
 
1.0 Icon.jpg
 
↑ compound visual receptors

↑ compound visual receptors

 

OUTTAKES: VLAD’S LITTLE SHOP OF HORRORS

when the machine had other plans

 
 
2020-07-27_21-26-21.gif
2020-07-12_17-32-09.gif
 
2020-07-27_20-42-39.gif
2020-07-28_00-46-44.gif
 
2020-07-28_11-32-19.gif
Kapture 2020-10-30 at 14.37.45.gif
 

A VISION OF THE FUTURE

This installation was developed and finalized mid-pandemic, and designed for a touchless environment. Viewers need only stand in front of one of the screens and change their facial expression to actively engage with the piece.

 
slide_mocks_3.jpg
 

ORIGINAL SCORE

Co-creating with AI software AIVA, our sound designer wove a short, darkly original composition, drawing inspiration from the pseudo-religious icon of the "ocular" used in the face reading prototype. They felt liturgic music composed by a computer, remixed and then performed by computers, would suit best.

slide_mocks_2.jpg
slide_mocks_1.jpg
 
 
 
 

Final prototype showcase

 

DISCLAIMER

This project is a collaboration between Master’s students at the Centre for Digital Media [CDM] and The Vancouver Art Gallery.
The prototype and all related content shown remains the IP of our client and the CDM.


 

Team 4149+i:

Courtney Clarkson, UX / UI Designer
Valentina Forte-Hernandez, Project Manager and Product Owner
Julia Read, UI Lead and Art Director
Vlad Ryzhov, Technical Lead
Patrick de Moss, Sound Design + original score
-
Larry Bafia, Director, CDM and Faculty Supervisor
Dr Patrick Parra Pennefather, Sound Supervisor