EdgeFusion: On-Device Text-to-Image Generation

arXiv:2404.11925v1 Announce Type: new
Abstract: The intensive computational burden of Stable Diffusion (SD) for text-to-image generation poses a significant hurdle for its practical application. To tackle this challenge, recent research focuses on methods to reduce sampling steps, such as Latent Consistency Model (LCM), and on employing architectural optimizations, including pruning and knowledge distillation. Diverging from existing approaches, we uniquely start with a compact SD variant, BK-SDM. We observe that directly applying LCM to BK-SDM with commonly used crawled datasets yields …

abstract application arxiv challenge computational consistency model cs.ai cs.cv cs.lg diffusion distillation image image generation knowledge practical pruning reduce research sampling stable diffusion text text-to-image type