.png){kind=logo}
BIO
Xavier Snelgrove (pronouns: he/him/his) (http://wxs.ca) is a computer scientist, researcher, and artist based in Toronto, Canada. He is a partner at Probably Studio (http://probablystudio.com), working with computer vision and machine learning for creative and biomedical applications. He is also Creative Technologist in Residence at the BMO Lab for Creative Research in the Arts, Performance, Emerging Technologies and AI at the University of Toronto. Previously he was a cofounder of an AI startup Whirlscape, and has been a researcher at places such as Element AI and Disney Research. Xavier is particularly interested in creative applications of AI research, and how the arts invite us to approach computation from new perspectives and ask different questions. He maintains a creative practice and regularly organizes workshops (https://genarthackparty.com) and galleries on computation and art, including at ICCV and ECCV, two premier international computer vision conferences (http://computervisionart.com).
Mattie Tesfaldet (pronouns: they/them) is a computer vision and machine learning researcher, artist, and DJ based in Montréal, Canada. They are pursuing their PhD at McGill University and Mila researching generative models for visual content creation, specifically, looking for novel and interesting ways images and videos can be represented with neural networks. Their work has been published at top international computer vision and machine learning conferences such as CVPR, NeurIPS, ICCV, and ECCV. Outside of academia, they like to apply their research with the aim of exploring the intersection of human creativity and artificial intelligence. Particularly, developing new AI-based mediums for communication, expression, and sharing of visual imagery. You can find them on instagram @mattierialgirl.
Video Loop - 1920 x 1920 px, 60 fps, 27’’
Latent Space Interpolation is a creation from the machine learning research practice of Mattie Tesfaldet and Xavier Snelgrove. These two artist-researchers, one based in Montreal, the other in Toronto collaborate regularly to investigate how algorithms can understand, represent, and produce imagery. Their work requires bidirectional translation between aesthetic and technical domains, with creative impulses given form in mathematical rigour.
For their scientific article entitled Fourier-CPPNs for Image Synthesis the artists invented a new algorithm to generate images based on recent artificial intelligence techniques. In Latent Space Interpolation, this algorithm moves between its idea of each of the artists' portraits via the uncharted territory in between. Of the infinite ways to transition between these images, the route taken hints at the algorithm's normally invisible internal structures and biases, providing new intuitions on its capacities, and new aesthetic directions to explore.