SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2108.00946 · 2021

StyleGAN-NADA: CLIP-Guided Domain Adaptation of Image Generators

arXiv · PDF · Open in the Atlas

Code that ran

We lifted 10 functions out of this paper's own repositories and ran 4 of them in a sandbox. "Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results.

RepositoryRoleRan
rinongal/StyleGAN-nada canonical 4 of 10
FunctionStatusWhere it lives
convert_conv Ran rinongal/StyleGAN-nada/convert_weight.py
code served (permissive licence) · get_code("7fe245a7216ae67b")
convert_modconv Ran rinongal/StyleGAN-nada/convert_weight.py
code served (permissive licence) · get_code("1c7cd5abc48bcf75")
convert_torgb Ran rinongal/StyleGAN-nada/convert_weight.py
code served (permissive licence) · get_code("96b279b9ef818d94")
make_kernel Ran rinongal/StyleGAN-nada/ZSSGAN/model/sg2_model.py
code served (permissive licence) · get_code("6f65e378a4313f87")
ask_yes_no Not yet run rinongal/StyleGAN-nada/ZSSGAN/dnnlib/util.py
code served (permissive licence) · get_code("9d31d2c4cd16bb2d")
duplicate_latent Not yet run rinongal/StyleGAN-nada/ZSSGAN/generate_videos.py
code served (permissive licence) · get_code("323317029c8b57cb")
format_time Not yet run rinongal/StyleGAN-nada/ZSSGAN/dnnlib/util.py
code served (permissive licence) · get_code("053fc534bc6bb989")
format_time_brief Not yet run rinongal/StyleGAN-nada/ZSSGAN/dnnlib/util.py
code served (permissive licence) · get_code("77f4aa0649e7f404")
interpolate_forward_backward Not yet run rinongal/StyleGAN-nada/ZSSGAN/generate_videos.py
code served (permissive licence) · get_code("0929a327abb07197")
project_code Not yet run rinongal/StyleGAN-nada/ZSSGAN/generate_videos.py
code served (permissive licence) · get_code("53b619216cd18d2b")

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

Can a generative model be trained to produce images from a specific domain, guided by a text prompt only, without seeing any image? In other words: can an image generator be trained "blindly"? Leveraging the semantic power of large scale Contrastive-Language-Image-Pre-training (CLIP) models, we present a text-driven method that allows shifting a generative model to new domains, without having to collect even a single image. We show that through natural language prompts and a few minutes of training, our method can adapt a generator across a multitude of domains characterized by diverse styles and shapes. Notably, many of these modifications would be difficult or outright impossible to reach with existing methods. We conduct an extensive set of experiments and comparisons across a wide range of domains. These demonstrate the effectiveness of our approach and show that our shifted models maintain the latent-space properties that make generative models appealing for downstream tasks.

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2108.00946")
get_code_for_paper("2108.00946")
have("2108.00946")

Connect an agent — have() is free.