Imbalance-Robust and Sampling-Efficient Continuous Conditional GANs via Adaptive Vicinity and Auxiliary Regularization

TL;DR

This paper introduces CcGAN-AVAR, a novel conditional GAN framework that addresses data imbalance and sampling efficiency issues in continuous generative modeling, achieving state-of-the-art results.

Contribution

The paper proposes adaptive vicinity and multi-task discriminator components to improve data balance handling and introduces a one-step generator for faster sampling.

Findings

Achieves 30x-2000x faster inference than CCDM.

Outperforms existing methods in generation quality across multiple datasets.

Effectively handles data imbalance with adaptive mechanisms.

Abstract

Recent advances in conditional generative modeling have introduced Continuous conditional Generative Adversarial Network (CcGAN) and Continuous Conditional Diffusion Model (CCDM) for estimating high-dimensional data distributions conditioned on scalar, continuous regression labels (e.g., angles, ages, or temperatures). However, these approaches face fundamental limitations: CcGAN suffers from data imbalance due to fixed-size vicinity constraints, while CCDM requires computationally expensive iterative sampling. To address these issues, we propose CcGAN-AVAR, an enhanced CcGAN framework featuring (1) two novel components for handling data imbalance - an adaptive vicinity mechanism that dynamically adjusts vicinity size and a multi-task discriminator that enhances generator training through auxiliary regression and density ratio estimation - and (2) the GAN framework's native one-step generator, enable 30x-2000x faster inference than CCDM. Extensive experiments on four benchmark datasets (64x64 to 256x256 resolution) across eleven challenging settings demonstrate that CcGAN-AVAR achieves state-of-the-art generation quality while maintaining sampling efficiency.