English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Classifier guidance -- using the gradients of an image classifier to steerthe generations of a diffusion model -- has the potential to dramaticallyexpand the creative control over image generation and editing. However,currently classifier guidance requires either training new noise-aware modelsto obtain accurate gradients or using a one-step denoising approximation of thefinal generation, which leads to misaligned gradients and sub-optimal control.We highlight this approximation's shortcomings and propose a novel guidancemethod: Direct Optimization of Diffusion Latents (DOODL), which enablesplug-and-play guidance by optimizing diffusion latents w.r.t. the gradients ofa pre-trained classifier on the true generated pixels, using an invertiblediffusion process to achieve memory-efficient backpropagation. Showcasing thepotential of more precise guidance, DOODL outperforms one-step classifierguidance on computational and human evaluation metrics across different formsof guidance: using CLIP guidance to improve generations of complex prompts fromDrawBench, using fine-grained visual classifiers to expand the vocabulary ofStable Diffusion, enabling image-conditioned generation with a CLIP visualencoder, and improving image aesthetics using an aesthetic scoring network.Code at https://github.com/salesforce/DOODL.

End-to-End Diffusion Latent Optimization Improves Classifier Guidance