Computational and Historical Analysis of Hands and Gestures in Early Modern Art- Valentine Bernasconi - PhD Fellow


I am currently a PhD student in the field of Digital Visual Studies at the University of Zurich (UZH). Prior to that, I graduated from the Ecole Polytechnique Fédérale de Lausanne (EPFL) with a M.Eng. in Digital Humanities. I also have a previous M.Sc. in Multimedia Design and 3D Technologies from Brunel University London and a B.Sc. in Computer Science from the University of Fribourg, where I received the JAACS award for best Bachelor thesis. The latter was part of an interdisciplinary collaboration with dancers from the University of the Art of Biel (HKB) and engineers from the Haute Ecole d’Ingénierie et d’Architecture of Fribourg (HEIA-FR).

My main research interest lies in between art and technology. Thanks to a minor in art history at the University of Fribourg, I was able to develop a critical approach to artistic content alongside an education in computer science. This interdisciplinary knowledge allowed me to work on a collaborative project called Nautilus. I there investigated the rise of digital media in performing arts from an art history point of view and contributed to the project with an interactive visualization. I further explored the world of augmented reality and 3D experiences with a master thesis called Interactive live 3D visualization.  Finally, I had the opportunity to delve into the world of design at the EPFL+ECAL Lab with a master thesis Valorization of visual heritage through A.I. algorithm trained on curated content. The curated content, the poster collection from the Museum für Gestaltung Zürich, allowed me to approach the field from both an art history and artistic perspective and to explore a diversity of computer vision tools. The research lead to the creation of a machine learning model to reproduce the creative thinking of the designer.

The core of my current research is at the intersection of computer vision and art history and is supervised by Prof. Dr. Tristan Weddigen and Dr. Leonardo Impett. I there wish to explore recurrent patterns in pictorial art and their evolution in time and place. This approach relies on the context of visual and cultural influence of artworks and intend to produce new exploring tools to the field of art history.

PhD Project - Lineage of characteristic figures and patterns in pictorial art. On hand and hand gestures in the early Renaissance

Project Summary


Reproductions have an important place in art history and the notion of a context of visual and cultural influence is non-negligible. These reproductions do not only involve artworks in their entirety, but also patterns and replicas, smaller parts, or similar representations that can be found from one painting to another. The Italian Renaissance seems to be a rich field of exploration for such replicas and patterns, especially regarding their expansion and evolution.

As an entry point to the vastness of the present research’s subject, hand and hand gestures are explored. Primary research has shown that the early Renaissance represents an interesting shift in the use of hand positions and gestures, from highly symbolic poses to their use in new contexts, with a general evolution towards more natural movements and descriptive gestures at the service of the narration. The phenomenon called transferred conventional gestures, was outlined in the work of Barasch on Giotto’s artworks.

To what extent are symbolic gestures used in new iconographic themes? How does the phenomenon evolve over time? Who were the initiators of such new representations and what were their possible influences? Dealing with these primary queries, the goal of the project is to analyze artistic material with the help of existing computer tools tailored to the medium. The main approach is to extract hands and gestures from a collected dataset of artworks dating from the late Middle Ages until the 15th Century from the Fototeca of the Bibliotheca Hertziana. Based on this extraction, unsupervised clustering of the hands will be performed in order to outline possible repetitive patterns and perform further analyses based on the available metadata. The latter might involve the exploration of a specific hand pose and its evolution in time and place, or the evolution of a representation in the context of a specific iconography.