Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

What Makes Synthetic Data Effective in Image Segmentation

About

Driven by rapid advances in large-scale generative models, synthetic data has emerged as a promising solution for visual understanding. While modern diffusion models achieve remarkable photorealistic image synthesis, their potential in complex visual segmentation tasks remains underexplored. In this work, we conduct a systematic analysis of synthetic images from state-of-the-art diffusion models to uncover the factors governing their utility. In particular, synthetic images characterized by dense scene composition and fine instance fidelity demonstrate distinctive benefits, yielding significantly more discriminative spatial representations. Building on these insights, we propose SENSE, a unified framework that leverages flexible and scalable synthetic data to substantially enhance segmentation performance. Notably, SENSE is model-agnostic, compatible with diverse architectures (e.g., DPT and Mask2Former), and scales effectively across models with varying parameter capacities. Extensive experiments on Cityscapes, COCO, and ADE20K validate the effectiveness and generalization capability of our approach. Code is available at https://github.com/zhang0jhon/SENSE.

Jinjin Zhang, Xiefan Guo, Yizhou Jin, Nan Zhou, Di Huang• 2026

Related benchmarks

TaskDatasetResultRank
Semantic segmentationADE20K (val)
mIoU58.27
3069
Semantic segmentationADE20K
mIoU (SS)59.09
21
Showing 2 of 2 rows

Other info

Follow for update