Research map / Image generation / Editing and style transfer
Unsupervised Image-to-Image Translation Networks
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Use UNIT instead of DTN | Unsupervised domain adaptation · SVHN→MNIST | classification accuracy: 0.8488 -> 0.9053 | Helped |
| Use UNIT instead of CoGAN | Unsupervised domain adaptation · MNIST→USPS | classification accuracy: 0.9565 -> 0.9597 | Helped |
| Use UNIT instead of CoGAN | Unsupervised domain adaptation · USPS→MNIST | classification accuracy: 0.9315 -> 0.9358 | Helped |
| Add the weight-sharing constraint and reconstruction streams (ablation) | CycleGAN-like model · Map satellite→map test set | average pixel accuracy: 0.569 -> 0.600 | Helped |
| Add cycle-consistency to the weight-sharing model (ablation) | UNIT · Map satellite→map test set | average pixel accuracy: 0.568 -> 0.600 | Helped |
| Vary the number of weight-sharing layers from 1 to 4 (ablation) | UNIT · Map satellite→map test set | translation accuracy | No clear effect |
| Use the shallowest discriminator architecture (ablation) | UNIT · Map satellite→map test set | translation accuracy | Hurt |
| Increase the negative log-likelihood term weights (ablation) | UNIT · Map satellite→map test set | translation accuracy | Helped |
| Set KL-term weights to 0.1 (ablation) | UNIT · Map satellite→map test set | translation accuracy | Helped |
| Remove cycle-consistency (ablation) | UNIT for domain adaptation · Domain adaptation | classification accuracy | No clear effect |
Closest papers by meaning
- TraVeLGAN: Image-To-Image Translation by Transformation Vector Learning (CVPR 2019 · similarity 0.93)
- Unsupervised Image-to-Image Translation With Generative Prior (CVPR 2022 · similarity 0.93)
- Image to Image Translation for Domain Adaptation (CVPR 2018 · similarity 0.91)
- Latent Filter Scaling for Multimodal Unsupervised Image-To-Image Translation (CVPR 2019 · similarity 0.90)
- Unsupervised Image-to-Image Translation Using Domain-Specific Variational Information Bound (NeurIPS 2018 · similarity 0.90)
- Rethinking the Paradigm of Content Constraints in Unpaired Image-to-Image Translation (AAAI 2024 · similarity 0.88)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.