EqGAN: Feature Equalization Fusion for Few-shot Image Generation
This work addresses the challenge of generating high-quality and diverse images with limited data, which is crucial for applications in computer vision, but it appears incremental as it builds on existing fusion-based methods.
The paper tackles the problem of poor quality and diversity in few-shot image generation by proposing EqGAN, a feature equalization fusion method that improves FID scores by up to 32.7% and LPIPS scores by up to 4.19% on public datasets.
Due to the absence of fine structure and texture information, existing fusion-based few-shot image generation methods suffer from unsatisfactory generation quality and diversity. To address this problem, we propose a novel feature Equalization fusion Generative Adversarial Network (EqGAN) for few-shot image generation. Unlike existing fusion strategies that rely on either deep features or local representations, we design two separate branches to fuse structures and textures by disentangling encoded features into shallow and deep contents. To refine image contents at all feature levels, we equalize the fused structure and texture semantics at different scales and supplement the decoder with richer information by skip connections. Since the fused structures and textures may be inconsistent with each other, we devise a consistent equalization loss between the equalized features and the intermediate output of the decoder to further align the semantics. Comprehensive experiments on three public datasets demonstrate that, EqGAN not only significantly improves generation performance with FID score (by up to 32.7%) and LPIPS score (by up to 4.19%), but also outperforms the state-of-the-arts in terms of accuracy (by up to 1.97%) for downstream classification tasks.