AI RESEARCH
OpenVision 3: A Family of Unified Visual Encoder for Both Understanding and Generation
arXiv CS.AI
•
ArXi:2601.15369v2 Announce Type: replace-cross This paper presents a family of advanced vision encoder, named OpenVision 3, that learns a single, unified visual representation that can serve both image understanding and image generation. Our core architecture is simple: we feed VAE-compressed image latents to a ViT encoder and train its output to two complementary roles. First, the encoder output is passed to the ViT-VAE decoder to reconstruct the original image, encouraging the representation to capture generative structure.