DISTIME DISPLACE

VADIM EPSTEIN (RUSSIA)

Python/Tensorflow/PyTorch, custom neural networks & models; animating-landscape

6:00 mins

February 2020

 

Usually, dystopian visions depict urban disasters as physical demolitions [breaking walls, falling towers, all that jazz..]. Instead, ‘Distime Displace’ visually explores the allegory of transformation, which reminds that apocalypse is not the end, but a route. Enjoy the trip…

 

A few still frames were taken from the top time-lapse cityscapes, working as a start point ["You are here"]. A neural network has been trained to reproduce the motion of the light and greenery in forests - effectively bringing the natural driving forces into urban picture. Then, the frames have been animated by the network with a very slow pace, impressing the conversion of cell concrete into misorganic entities.