1,101 followers
New Research: Orthogonality and graph divergence losses promote disentanglement in generative models https://t.co/vTlIsIg1hs #computerscience
New Research: Orthogonality and graph divergence losses promote disentanglement in generative models https://t.co/vTlIsIg1hs #computerscience
RT @pturaga1: Controllable disentanglement feasible by adding geometry-based loss functions in latents and outputs. Below we disentangle by…
Controllable disentanglement feasible by adding geometry-based loss functions in latents and outputs. Below we disentangle by size, color, and shape. Exciting! https://t.co/qgk1zlnEBP @Ankita1Shukla @visarASU @GautamDasarathy @Rajhans_0 @pouria018 Rishi