DM-Tune: Quantizing Diffusion Models with Mixture-of-Gaussian Guided Noise Tuning

Diffusion models have become essential generative tools for tasks such as image generation, video creation, and inpainting, but their high computational and memory demands pose challenges for efficient deployment. Contrary to the traditional belief that full-precision computation ensures optimal ima...

Ausführliche Beschreibung

Bibliographische Detailangaben

Veröffentlicht in:	2025 62nd ACM/IEEE Design Automation Conference (DAC) S. 1 - 7
Hauptverfasser:	Haghi, Pouya, Falahati, Ali, Azad, Zahra, Wu, Chunshu, Song, Ruibing, Liu, Chuan, Li, Ang, Geng, Tong
Format:	Tagungsbericht
Sprache:	Englisch
Veröffentlicht:	IEEE 22.06.2025
Schlagworte:	Diffusion models Diversity reception GPU Graphics processing units Image quality Image synthesis Kernel Mixed-precision Noise Quantization Quantization (signal) Runtime Text to image
Online-Zugang:	Volltext
Tags:	Tag hinzufügen Keine Tags, Fügen Sie den ersten Tag hinzu!