MosaicFusion: diffusion models as data augmenters for large vocabulary instance segmentation

• Main Authors: Xie, Jiahao, Li, Wei, Li, Xiangtai, Liu, Ziwei, Ong, Yew Soon, Loy, Chen Change
• Other Authors: College of Computing and Data Science
• Format: Article
• Language: English
• Published: 2025
• Subjects: Computer and Information Science; Text-to-image diffusion models; Long tail
• Online Access: https://hdl.handle.net/10356/182213
• Institution: Nanyang Technological University

We present MosaicFusion, a simple yet effective diffusion-based data augmentation approach for large vocabulary instance segmentation. Our method is training-free and does not rely on any label supervision. Two key designs enable us to employ an off-the-shelf text-to-image diffusion model as a useful dataset generator for object instances and mask annotations. First, we divide an image canvas into several regions and perform a single round of diffusion process to generate multiple instances simultaneously, conditioning on different text prompts. Second, we obtain corresponding instance masks by aggregating cross-attention maps associated with object prompts across layers and diffusion time steps, followed by simple thresholding and edge-aware refinement processing. Without bells and whistles, our MosaicFusion can produce a significant amount of synthetic labeled data for both rare and novel categories. Experimental results on the challenging LVIS long-tailed and open-vocabulary benchmarks demonstrate that MosaicFusion can significantly improve the performance of existing instance segmentation models, especially for rare and novel categories. Code: https://github.com/Jiahao000/MosaicFusion.