Ash Koosha & Dalena Tran tell stories with poor image-making algorithms in their video ‘Incomplete’

, 4 June 2021
video

Ash Koosha and Dalena Tran are releasing ‘Incomplete’, a new video work premiering on AQNB today, June 4, 2021. The video resulted from Tran’s research into open-source imaging software, 3D modeling, and animation. 

Enmeshed throughout ‘Incomplete’, bare constructions of animals and landscapes slowly glitch into loose configurations of human infrastructure as the video goes on with Koosha’s crooning overtones. About her inspiration to use Generative Adversarial Networks (GANs), Tran explains in the press release: 

“Many GAN-related works we see today focus on the interpolation of latent space and the fluid changes from one subject to another. With this work, I wanted to evoke an expression of storytelling using machine learning to expose the influx of poor images; providing the aesthetic and subjective quality that brings the attention back into the process itself.”**

Ash Koosha and Dalena Tran’s ‘Incomplete’ is out today, June 4, 2021.