Rendered at 17:59:17 GMT+0000 (Coordinated Universal Time) with Wasmer Edge.
vighneshb 6 days ago [-]
Hi
In our latest paper we shoa that a GAN loss (used by almost all latent diffusion models) to train their autoencoders is not required and instead can be replaced with a diffusion loss. Our auto-encoder is trained end-to-end and achieves higher compression and better generation quality.
I am excited to share it with you. Let me know what you think.
it seems to be another autoencoder(autoregressive) + diffusion.
vighneshb 6 days ago [-]
This is very interesting. Unlike us (who focus on the decoder) they focus on changing the representation itself so that they can achieve better generation. Thanks for the link.
billconan 6 days ago [-]
they use autoencoder/autoregressive model to predict the big picture, and diffusion for the details, similar to yours.
In our latest paper we shoa that a GAN loss (used by almost all latent diffusion models) to train their autoencoders is not required and instead can be replaced with a diffusion loss. Our auto-encoder is trained end-to-end and achieves higher compression and better generation quality.
I am excited to share it with you. Let me know what you think.
Cheers
it seems to be another autoencoder(autoregressive) + diffusion.
The difference is they use discrete tokens.