Research map / Image generation / Editing and style transfer
Unsupervised Attention-guided Image-to-Image Translation
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 |
|---|---|---|---|
| add unsupervised attention-guided image translation | CycleGAN-style unpaired image translation · Apple to Orange; Orange to Apple; Zebra to Horse; Horse to Zebra; Lion to Tiger; Tiger to Lion | KID ×100: 8.48 $\pm$ 0.53 -> 6.44 $\pm$ 0.69; 9.82 $\pm$ 0.51 -> 5.32 $\pm$ 0.48; 11.44 $\pm$ 0.38 -> 8.87 $\pm$ 0.26; 10.25 $\pm$ | Helped |
| add cycle-consistency loss (ablation) | attention-guided image translation · Zebra to Horse; Horse to Zebra | KID ×100: 64.55 $\pm$ 0.34 -> 8.87 $\pm$ 0.26; 41.48 $\pm$ 0.34 -> 6.93 $\pm$ 0.27 | Helped |
| compute a new attention map on the inverse mapping instead of reusing the source attention (ablation) | attention-guided image translation · Zebra to Horse; Horse to Zebra | KID ×100: 9.46 $\pm$ 0.38 -> 8.87 $\pm$ 0.26; 7.79 $\pm$ 0.23 -> 6.93 $\pm$ 0.27 | Helped |
| use both attention networks instead of only A_S (ablation) | attention-guided image translation · Zebra to Horse; Horse to Zebra | KID ×100: 10.90 $\pm$ 0.25 -> 8.87 $\pm$ 0.26; 7.62 $\pm$ 0.25 -> 6.93 $\pm$ 0.27 | Helped |
| use both attention networks instead of only A_T (ablation) | attention-guided image translation · Zebra to Horse; Horse to Zebra | KID ×100: 9.30 $\pm$ 0.45 -> 8.87 $\pm$ 0.26; 7.80 $\pm$ 0.21 -> 6.93 $\pm$ 0.27 | Helped |
| use an attention-guided discriminator instead of a whole-image discriminator (ablation) | attention-guided image translation · Zebra to Horse; Horse to Zebra | KID ×100: 9.26 $\pm$ 0.22 -> 8.87 $\pm$ 0.26; 7.77 $\pm$ 0.35 -> 6.93 $\pm$ 0.27 | Helped |
| stop attention-network training and use an attention-guided discriminator after the initial stage (ablation) | attention-guided image translation · Zebra to Horse; Horse to Zebra | KID ×100: 9.86 $\pm$ 0.32 -> 8.87 $\pm$ 0.26; 8.28 $\pm$ 0.34 -> 6.93 $\pm$ 0.27 | Helped |
Closest papers by meaning
- U-GAT-IT: Unsupervised Generative Attentional Networks with Adaptive Layer-Instance Normalization for Image-to-Image Translation (ICLR 2020 · similarity 0.90)
- Unsupervised Image-to-Image Translation Networks (NeurIPS 2017 · similarity 0.90)
- Unsupervised Image-to-Image Translation With Generative Prior (CVPR 2022 · similarity 0.89)
- LookHere: Vision Transformers with Directed Attention Generalize and Extrapolate (NeurIPS 2024 · similarity 0.87)
- Improving Adversarial Transferability on Vision Transformers via Forward Propagation Refinement (CVPR 2025 · similarity 0.85)
- On the Surprising Effectiveness of Attention Transfer for Vision Transformers (NeurIPS 2024 · similarity 0.85)
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.