17th IAFF Animator will take place in Poznań June 25-30, 2024 at the Kino Muza and other locations in the city center.
This year, the Animator Festival is looking towards artificial intelligence and the works that began to be created with its participation. In the section titled ARTIFICIAL INTUITION, from over a hundred submitted works, we will see films focusing on the impact that AI driven tools have on creativity and close relations. The Festival will also be an open platform for discussion about changes in understanding of human identity and its invention.
Witnessing the development of artificial intelligence tools and its growing impact on our environment, we decided to invite AI to cooperate in this year's visual identification of the Festival. The challenge of creation with the participation of artificial intelligence has been taken up by Jakub de Barbaro who is an independent designer, cooperating so far with institutions such as: the Museum of Modern Art in Warsaw, Sternberg Press in Berlin, Mousse Publishing in Milan and T4ex in New York. In his practice, De Barbaro tries to focus on the non-commercial functions of design, and treats his work as a tool that enhances sensitivity and shapes human attitude as well as social behavior. The aspect of digital ecology is important to him, which is why he constantly cooperates with the Digital Ecology Institute.
Jakub de Barbaro says this about this year's identification of the Festival:
The idea of using artificial intelligence is related to this year's festival tagline: "Artificial Empathy". The festival's identification design consists of several dozen generated images, trans-species animal figures, hiding the shape of the festival's sign - letter (a.). The responsible use of artificial intelligence in design requires transparent access to source databases, information processors and models used. I used Artbreedr.com software to work. The biological model was created by Xander Steenbrugge. The furry image model was built by arfa.dev. The generated furry portraits were trained using nVidia's StyleGAN2 architecture. The training data set consisted of approximately 55,000 SFW images from e621.net processed using a custom YOLOv3 network. The network was trained for 33 days (3.2 million iterations) on the TPUv3-32 device. The final images were upscaled to 2x resolution and denoised using Waifu2x.
Get ready for Animator!
See you on June 25 - 30, 2024 at Kino Muza and other locations in the center of Poznań. The full program will be announced in early June. Stay tuned for ticket sales and accreditations.
