Image to Image Translation

Image-to-Image Translation transforms images from one visual domain to another whilst preserving content structure, converting between image modalities such as sketch-to-photo, day-to-night, satellite-to-map, or style transfer between artistic styles.

Semantic Classification

Content

• Image-to-Image Translation transforms images from one visual domain to another whilst preserving content structure, converting between image modalities such as sketch-to-photo, day-to-night, satellite-to-map, or style transfer between artistic styles. Image translation models (Pix2Pix, CycleGAN, StyleGAN) employ conditional generation and adversarial learning to learn mappings between paired or unpaired image domains.

DOING Images

• Landscape renders using custom Stable Diffusion LoRAs (for example, JJ’s Landscape Render, StreetScape and Ecology Park models on Civitai)

• General tools such as Midjourney, Stable Diffusion and DALL·E 3

  

  

• Image Generation

  • Image processing from drawings

• Experimental generative video platforms (Runway ML Gen-2, PromeAI) for short concept clips

• Sustainable Home Visuals AI

• JJ’s Landscape Render - XL v1.0 | Stable Diffusion XL LoRA | Civitai

• JJs StreetScape - XL v1.0 | Stable Diffusion XL LoRA | Civitai

• JJs Ecology Park - XL v1.0 | Stable Diffusion XL LoRA | Civitai

  Core Characteristics

• Domain Transfer: Mapping between visual domains

• Structure Preservation: Maintaining spatial and semantic structure

• Paired or Unpaired: Supervised (Pix2Pix) or unsupervised (CycleGAN) learning

• Bidirectional Translation: Cycle-consistency for unpaired domains

• Multi-Modal Applications: Medical imaging, satellite imagery, artistic style

  Relationships

• Subclass: Computer Vision, Image Generation

• Related: Style Transfer, Generative Adversarial Network

• Models: Pix2Pix, CycleGAN, UNIT, MUNIT, StarGAN

• Applications: Medical Imaging, Remote Sensing, Creative Tools

  Key Literature

  1. Isola, P., et al. (2017). “Image-to-image translation with conditional adversarial networks.” CVPR, 1125-1134.

  2. Zhu, J. Y., et al. (2017). “Unpaired image-to-image translation using cycle-consistent adversarial networks.” ICCV, 2223-2232.

  See Also

• Image Generation

• Style Transfer

• Generative Adversarial Network

Provenance