An experimental text-to-image generation system that learns a shared latent representation between vision and language by mapping semantic text embeddings into an image autoencoder’s latent space and reconstructing images from these aligned representations.
AKASH-KUMAR-BARPANDA/Text_to_image_through_latent_representation: An experimental text-to-image generation system that learns a shared latent representation between vision and language by mapping semantic text embeddings into an image autoencoder’s latent space and reconstructing images from these aligned representations.
