English

Latent Space Synergy: Text-Guided Data Augmentation for Direct Diffusion Biomedical Segmentation

Image and Video Processing 2025-07-22 v1 Artificial Intelligence Computer Vision and Pattern Recognition

Abstract

Medical image segmentation suffers from data scarcity, particularly in polyp detection where annotation requires specialized expertise. We present SynDiff, a framework combining text-guided synthetic data generation with efficient diffusion-based segmentation. Our approach employs latent diffusion models to generate clinically realistic synthetic polyps through text-conditioned inpainting, augmenting limited training data with semantically diverse samples. Unlike traditional diffusion methods requiring iterative denoising, we introduce direct latent estimation enabling single-step inference with T x computational speedup. On CVC-ClinicDB, SynDiff achieves 96.0% Dice and 92.9% IoU while maintaining real-time capability suitable for clinical deployment. The framework demonstrates that controlled synthetic augmentation improves segmentation robustness without distribution shift. SynDiff bridges the gap between data-hungry deep learning models and clinical constraints, offering an efficient solution for deployment in resourcelimited medical settings.

Keywords

Cite

@article{arxiv.2507.15361,
  title  = {Latent Space Synergy: Text-Guided Data Augmentation for Direct Diffusion Biomedical Segmentation},
  author = {Muhammad Aqeel and Maham Nazir and Zanxi Ruan and Francesco Setti},
  journal= {arXiv preprint arXiv:2507.15361},
  year   = {2025}
}

Comments

Accepted to CVGMMI Workshop at ICIAP 2025

R2 v1 2026-07-01T04:10:44.876Z