Daniel Berio

Research

Home  > Research

In the last few years, I have been researching ways to computationally reproduce the aesthetics of graffiti art. I recently defended my PhD thesis on this topic, which is titled AutoGraff: towards a computational understanding of graffiti writing and related art forms.

You can download the thesis document here. Supplemental material and videos for each chapter of the thesis can be accessed here.

The thesis covers techniques inspired by principles of human motor control to design and generate expressive curves, methods to rapidly design sinthetic graffiti in different styles, and methods for graffiti stylisation of text and fonts. Many of the graphics in this website are experiments that relate to the technologies that I developed in the thesis.

You can find more info, watch videos and download publications at the Autograff project website or on Researchgate.

The project was funded by EPSRC and in a later stage supported by Adobe Research

Talks

In addition to various published paper presentations, I have given a number of talks on the topic of the thesis, the following is a list of the most relevant:

  • May 2019,Frederic Fol Leymarie and Daniel Berio, Movement Computing to Model a Class of Visual Art Productions Towards understanding human creative activities of writing, calligraphy, graffiti and beyond.
    • Konstanz University, Konstanz.
  • December 2018, Daniel Berio, Keynote: Computational models of graffiti form.
  • May 2018, Daniel Berio and Frederic Fol Leymarie, Seminar: Doing research on art, human and artificial intelligence, intersecting.
    • McGill University, Computer graphics department, Montreal.
  • July 2017, Daniel Berio, Graffiti synthesis, a motion centric approach.
    • Creative AI meetup, London

Publications

  • Berio, Fol Leymarie & Calinon, Interactive Generation of Calligraphic Trajectories from Gaussian Mixtures, Mixture Models and Applications, 23-38 (2019).
  • Berio, Asente, Echevarria & Fol Leymarie, Sketching and Layering Graffiti Primitives, 51-59, in: 8th ACM/Eurographics Expressive Symposium on Computational Aesthetics (2019)
  • Berio, Leymarie & Plamondon, Kinematic Reconstruction of Calligraphic Traces from Shape Features, 762-767, in: Proceedings of the International Conference on Pattern Recognition and Artificial Intelligence
  • Berio, Calinon & Fol Leymarie, Generating Calligraphic Trajectories with Model Predictive Control, in: Proceedings of Graphics Interface, edited by Canadian Human-Computer Communications Society (2017)
  • Berio, Akten, Fol Leymarie, Grierson & Plamondon, Calligraphic Stylisation Learning with a Physiologically Plausible Model of Movement and Recurrent Neural Networks, in: Proc. of 4th Int’l Conf. on Movement Computing (MOCO) (2017)
  • Berio, Fol Leymarie & Plamondon, Computer Aided Design of Handwriting Trajectories with the Kinematic Theory of Rapid Human Movements, in: 18th Biennial Conference of the International Graphonomics Society (2017).
  • Berio, Calinon & Fol Leymarie, Learning dynamic graffiti strokes with a compliant robot, in in: Proc. IEEE/RSJ Intl Conf. on Intelligent Robots and Systems (IROS) (2016)
  • Berio, Calinon & Leymarie, Dynamic Graffiti Stylisation with Stochastic Optimal Control, in in: Proceedings of the 4th International Conference on Movement Computing (2017)
  • Berio, Leymarie & Plamondon, Expressive Curve Editing with the Sigma Lognormal Model, 33-36, in in: Proceedings of the 39th Annual European Association for Computer Graphics Conference: Short Papers (2018)
  • Berio & Leymarie, Computational Models for the Analysis and Synthesis of Graffiti Tag Strokes, 35-47, in in: Computational Aesthetics (2015)