# Efficient Covariance-Aware Vicinal Latent Diffusion for Continuous Conditional Image Generation

**Authors:** Xin Ding, Shuo Huang, Zhenglong Ding, Sen Zhang, Nenglun Chen, Peibei Cao, Yongwei Wang, Fei Wu  
**Lab:** Nanjing University of Information Science and Technology · **Published:** 29 Oct 2026  
**Links:** [arXiv](https://arxiv.org/abs/2610.39277) · [PDF](https://arxiv.org/pdf/2610.39277) · [This page](https://omen-bit.github.io/ai-paper-daily/paper/2610.39277.html)

## Abstract

Continuous Conditional Generative Modeling (CCGM) aims to synthesize high-dimensional data conditioned on continuously valued scalar labels. However, existing continuous diffusion models face severe data sparsity under imbalanced and long-tailed label distributions, which can lead to visual artifacts and condition inconsistency. Moreover, inference-time classifier-free guidance and dense pixel-space integration introduce substantial sampling overhead, limiting practical deployment. To address these challenges, we propose CAViLD, an efficient latent diffusion framework for continuous conditional generation. CAViLD first constructs a purified condition-aware latent space through density-adaptive vicinal learning, auxiliary semantic regularization, and NIQE-guided negative data augmentation. It then internalizes the conditional guidance direction through an EMA teacher, enabling single-branch conditional inference, and introduces covariance-aware progressive distillation for label-dependent probability-flow ODEs, supporting sampling down to four NFEs. Experiments on RC-49, UTKFace, and Steering Angle demonstrate a favorable trade-off among generation fidelity, continuous condition consistency, and sampling efficiency.

---

Notes written by AI from the paper, numbers copied from the abstract. Check the paper before you rely on them. From AI Papers Daily.