Pierre Commoy and Gilles Blanchard, also known as Pierre et Gilles, are French artists and romantic partners. They have been producing works together since 1976, creating a world where painting and photography meet. Their art is peopled by their friends and family, anonymous and famous, who appear in sophisticated life-size sets the artists build in their studio. They meticulously apply paint to the photographs once printed on canvas. Accomplished image creators, Pierre and Gilles have built up an extraordinary contemporary iconography on the frontier between art history and popular culture.
Tagging
Pierre et Gilles have a surrealist art style, so tags like surreal, dreamlike, ethereal, vivid colors, vibrant, etc. help. Their photos also have a soft focus so soft focus is helpful. Using portrait of <whatever> also helps. These tags were pruned from the dataset so the model should capture the look by default, but YMMV.
Frames
The model was trained on their framed work as well. Use decorative frame, <type of frame> frame to frame the photo you generate. You can add decorative frame to the negatives if frames begin showing up.
Description
FAQ
Comments (11)
Iconic style and I'm very glad for your work and sharing this with us! Thank you!
thanks!
Love it, I remember discovering Pierre et Gille as a college student and being instantly enamored by their work.
do you know what? after all of your work to make this lora, I Believe in future you could make David LaChapelle, this lora is perfect, I cry to think about this.
Thanks for the vote of confidence. David LaChapelle's style is a bit intimidating, but I'm a fan of his as well! I'm thinking of trying Robert Mapplethorpe, Herb Ritts, and Bruce Weber sometime down the line too.
@razputin that will be black and white galore!! can't wait to see it!
@Los_Norelos that is a lot of black and white lol! I might have to space them out - good point. Thanks for the tip as well - I really appreciate it!
Hi! And thank you dearly for your great work!
Could you please remake it for Illustrious or NoobAI? Or share a training dataset? Please!
Sure - I can try. I didn't think anyone would be interested in an Illustrious/NoobAI version since they're more illustrative-based and not focused on realism/surrealism.







