Abstract
Can a generative model be trained to produce images from a specific domain, guided only by a text prompt, 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 infeasible to reach with existing methods. We…
Citation impact
467
total citations
- FWCI
- 44.65
- Percentile
- 100%
- References
- 32
Citations per year
Authors
6Topics & keywords
Topics
Keywords
- Computer science
- Generative grammar
- Generator (circuit theory)
- Generative model
- Set (abstract data type)
- Code (set theory)
- Image (mathematics)
- Artificial intelligence
UN Sustainable Development Goals
- Quality Education
No related works found for this paper.