1 paper · 1 filter
Shaoxiong Guo, Tianyi Du, Lijun Li +3
Unified Multimodal Understanding and Generation Models (UMMs) increasingly combine visual understanding and image generation within a single interactive workflow, making generated…