Sole Traverso earned her Bachelor’s and Master of Fine Arts degrees, specialising in painting, printmaking, and photography from the Berlin University of the Arts. Guided by her passion for computing, she pursued a second master’s degree, an MFA in Art and Technology Studies, from the School of the Art Institute of Chicago. Here, she specialised in 3D computer animation, virtual reality and programmed with OpenGL and C.
Upon her return to Germany, Sole lectured on multimedia and programming at Fachhochschule Potsdam. Subsequently, she began a career as a software engineer, dedicating more than a decade to small and large enterprises such as ifolor, Neofonie GmbH, T-Systems Multimedia, mobile.de/ebay, and Polaroid, where her research focused on computer vision-based projects and the interface between hardware and software.
This role awakened her curiosity to delve deeper into computer vision, inspiring her to pursue a third master’s degree, an MSc in Artificial Intelligence, from the University of Bath in the UK. While at Bath, Sole was introduced to Reinforcement Learning (RL) and sequential decision-making. She became interested in developing data-efficient and generalisable RL strategies, such as creating transferable skills for solving diverse tasks in a single environment. For her master’s dissertation, she explored Hierarchical Reinforcement Learning techniques based on spectral graph theory, focusing on skill generation using the eigenvectors of the graph Laplacian or those of the Successor Representation.
Today, Sole is a doctoral researcher at the Sprekeler lab and SCIoI, contributing to project 46, “Lockbox 2.0″.