Ivan Iovine

Ivan Iovine is an interaction designer and artist based in Frankfurt am Main. His artistic and media practices range from digital art to robotics and machine learning.Iovine uses emerging technologies to blend virtual and physical worlds, creating speculative scenarios that highlight challenges and issues in the use (or disuse) of these innovations and how these affect social contexts and paradigms.

He has exhibited his work in various international festivals and museums such as Lab30 (Augsburg – 2019), Piksel Festival (Bergen – 2020-2022), Die Digitale (Dusseldorf – 2021), Dortmunder U (Dortmund – 2020) and Run Run Shaw Creative Media Centre (Hong Kong – 2021). In 2016 he was invited by Arduino to attend the “Casa Jasmina” residency coordinated by writers and futurists Sterling and Tesanovic. He has written artistic and scientific publications for conferences like “Art Machines 2: International Symposium on Machine Learning and Art” (CityU Hong Kong – 2021) and “State of Interim” (HFBK Hamburg – 2022). Iovine worked for five years at Fraunhofer IGD in Darmstadt, where he developed geospatial applications in the fields of 3D graphics and virtual reality. Since May 2022, he leads the robotics lab at HfG Offenbach, where he teaches robotics, machine learning and computer vision.

ARTWORK

Theodore y Prototyping DIY smart robots with Arduino and Machine Learning
Installation

Theodore is an interactive installation capable to write sentimental letters based only on image content. The main idea underlying this project is to find out if machines are able to understand feel- ings, context and generate intimate and confidential letters that express human- like sentiments using only pictures as input. For this purpose, several machine learning frameworks are used in the field of facial expression detection, landscape recognition and text generation.

The in- stallation was exhibited at the JS Nation Festival in Amsterdam (2019), at LAB30 Media Art Festival in Augsburg (2019), and at Piksel Festival in Bergen (2020). In 2021 the work was published in the of- ficial journal proceedings of the scientific symposium “Art Machines 2: International Symposium on Machine Learning and Art”. The artwork was then exhibited for two weeks in the same year at the Run Run Shaw Creative Media Centre in Hong Kong.

The web application that the user interacts with was developed using MEAN Stack (Mongo, Express, Angular, Node). For image analysis and letter generation, computer vision technologies were used in conjunction with OpenAI. The images are analyzed using DenseCap, an object recognition model trained with the Visual Genome dataset.
The hardware part is developed with the help of an Arduino Mega. This manages the 32 Relays that turn on and off the 32 solenoids that automate the typewriter. The communication between software and hardware part is via serial communication.

Link a la obra
https://vimeo.com/346536739