# Decoupled architecture

VisionReasoner-7B is an image-text-to-text model that adopts a decoupled architecture and consists of a reasoning model and a segmentation model. It can interpret user intentions and generate pixel-level masks.

Seg-Zero-7B is an image segmentation model based on reasoning chain guidance, featuring a decoupled architecture that includes a reasoning model and a segmentation model. It achieves zero-shot generalization capabilities through GRPO reinforcement learning training.