Lida Zacharopoulou

Digital No-Man's : Do Androids Dream of Eclectic Steep?

Video Art ● AI Art ● Machine Learning ● Creative Coding ● Storytelling

2023

AI-generated images have nowadays become so realistic that distinguishing the real ones from the fakes is nearly impossible. Trained for days with massive data and GPUs scattered all around the world, they have crossed the borders of the Uncanny Valley into the hills of visual perfection.

What about those digital beings that do not have the resources to complete their training? These low-res synthetic creatures are imagined as the low-class citizens of the Humanoid Realm. They can't afford to train forever nor to be GPU nomads. They are stuck in the Uncanny Valley, always striving to convince the Discriminator they look real enough, always dreaming of climbing the steep hill of human likeness.

A short animated film exploring AI through AI, this project aims to shed light on the inner workings of AI algorithms, specifically those responsible for creating realistic human faces, like the GAN models. GAN systems (Generative Adversarial Networks) are machine learning algorithms comprising two neural networks: the Discriminator and the Generator. The core idea behind this model is a fascinating game played between these two entities: the Generator continuously produces images in an attempt to deceive the Discriminator into perceiving them as real rather than synthetic. In response, the Discriminator assesses each image's authenticity. This intriguing back-and-forth between the two networks is played out repeatedly, resulting in the production of increasingly realistic images.

The video depicts a fictional dialogue between the Discriminator and the synthetic creatures it evaluates. These creatures engage in a Sisyphean effort, an endless struggle to reach an elusive perfection, as they strive to deceive the Discriminator and rise above their pixelated origins.

Official Selections and Screenings:
Yeast Festival (Limassol CY, Feb 2024)
Interface Festival (Art-Kino Croatia, Rijeka HR, Dec 2023)
ICONA Festival 2023 (Corfu GR, Nov 2023)
Animation Marathon 2023 (Online, Nov 2023)
FINALIST, ROS Film Festival (Teatro Arniches, Alicante ES, Nov 2023)
Screen it ASFA (Athens School of Fine Arts GR, Nov 2023)
5th International Forum of Perfomance Art (Drama GR, Sep 2023)

Haenyeo Memories

AI Art ● Visual Arts

2023

Haenyeo Memories is a physical photoalbum of AI-generated photographs of the Haenyeo divers.

Haenyeo (sea women) are the female divers of Jeju Island in South Korea. Part of the world's intangible cultural heritage, the Haenyeo are free diving, often holding their breath for over 2 minutes, in order to collect sea urchins, squids, and other marine species. Since the 18th century, diving on the island is already a female-dominated profession: the Haenyeo become the basis of Jeju's economy, establishing semi-matriarchal family structures. Today, most of these divers are over 70 years old, while their numbers have decreased dramatically.

Using an AI system, a collection of old photos inspired by the Haenyeo was generated. In the system's attempt to synthesize these photos, traditional costumes are mixed with diving suits, spacesuits are decorated with shells, and strange sea creatures are transformed into alien beings, in a retrofuturistic photo album straight out of a sci-fi movie.

What inspired me about the culture of Haenyeo is that they contributed to the advancement of women’s status in the island, promoted environmental sustainability and community involvement in management of fishing practices. I think the story of the Haenyeo women offers a glimpse of how the history of greek diving islands could have been, a way out of the over-fishing practices of the Mediterranean.

The physicality of a mid-century Asian photo album and the post-processing of the printed photographs to look worn, creates a nostalgic feel but holds memories of a past that never existed. Coming from an island of the East Aegean, I have spent most summers of my life diving and snorkeling for hours. This work is an homage to our summer long dives and to our memories, the confusion of what we have seen and what we have imagined.

Exhibitions:
Apnea Contemporary Art Exhibition, (European Centre Athens GR, March 2023)

Blackbox AI

AI ● HCI ● Installation ● Research ● Gender bias

2022

Blackbox AI is an multidisciplinary art installation exploring gender bias in Artificial Intelligence. Are AI algorithms biased? Why are many women (self-)excluded from the STEAM fields and the discussions around AI?

Drawing from the definition of the black box as a complex system whose internal workings are hidden or not readily understood, the exhibition invites guests to enter a physical maze and walk through various aspects of gender bias, presented using storytelling and AI algorithms, such as facial recognition and text-to-image generators.

The idea for the project was developed during EthicAI=LABS 2022, where it won the first prize (Sofia, May 2022). The installation was presented in EthicAI=FORUM: How can AI be right? (Athens, November 2022), as well as during Breaking the Bias Summer School (Athens, July 2023). The research behind this project led to the development of the paper titled A Multimodal Installation Exploring Gender Bias in Artificial Intelligence, vpresented at the 25th International Conference on Human-Computer Interaction in Copenhagen in July 2023 and published by Springer.
Collaborative project together with Mihaela Dobreva, Tea Rukavina, Vivian Stamou and Anastasia Nefeli Vidaki.

Tools: Research, Python, OpenCV, DALL-E, Beamers, Fabric

Links: Conference Paper here.

Attractive Interactions

Generative Art ● Creative Coding ● Visuals ● Projection

2019

Molecules and people interact with attractive or repulsive forces, switching behavior between being drawn together to form groups and needing their space. This generative visuals piece consists of particles that randomly change their mind about feeling social or antisocial.
Selected as one of the top projects of ProgramaLaPlaza Festival 2019, it was projected on the facade of Medialab-Prado in Madrid during Fall-Winter 2019.

Tools: Processing (p5.js), Javascript

Links: You can run the project here to see the colors correctly (the video does not capture the true colours) as well as the differences of each repetition.
Technical details and code available on Github.

Sónar+D Innovation Challenge

Interactive Interface ● Game Development ● HCI ● Affective Computing ● Embodied Interaction

2019

Developed in the context of the Affective computing for learning through games challenge by Telefónica Alpha, where we were asked to create a game with an agent that detects affective states and delivers adaptive content in real-time.
This prototype for a music video game detects the excitement of the user and adjusts the difficulty of the game. Players must hit the correct drum pattern appearing on the screen, using either Arduino-based or MIDI controllers. A Bitalino sensor is capturing the user's heart rate, which adapts the tempo of the song in real time. The more excited the user, the faster the tempo gets.
Group project together with Héctor López Carral, Marco Musto, Alayna Hughes, and Lucas Lorenzo Peña developed during the 4 days of the Sónar 2019 festival.

Tools: Arduino, Sensors, Unity

Links: Presentation and documentation of the project here. See more about the 2019 challenge in this video by Sónar.

Penguin in the Desert

Real-Time Interaction ● Research ● Game Development ● Data Analysis ● Interactive Media

2019

A 2D platformer game to reserach the correlation between semantic proximity and object recognition. Players go through 4 levels (desert, beach, winter, jungle), collect various objects which are positively, negatively or not associated (random) with the background and are later asked to recognise them in a separate survey. Group project developed togeher with Anna Bellmunt, Paul Hine and Mila Sevilla Zhou.

Tools: Unity

Links: Presentation and documentation of the project here. You can play the game here.