COCONut-PanCap: Joint Panoptic Segmentation and Grounded Captions for Fine-Grained Understanding and Generation Paper • 2502.02589 • Published 12 days ago • 9
Democratizing Text-to-Image Masked Generative Models with Compact Text-Aware One-Dimensional Tokens Paper • 2501.07730 • Published Jan 13 • 16