Diversity-aware Channel Pruning for StyleGAN Compression

arXiv:2403.13548v1 Announce Type: new
Abstract: StyleGAN has shown remarkable performance in unconditional image generation. However, its high computational cost poses a significant challenge for practical applications. Although recent efforts have been made to compress StyleGAN while preserving its performance, existing compressed models still lag behind the original model, particularly in terms of sample diversity. To overcome this, we propose a novel channel pruning method that leverages varying sensitivities of channels to latent vectors, which is a key factor in sample …

abstract applications arxiv challenge compression computational cost cs.cv diversity however image image generation performance practical pruning sample terms type