Research map / Image generation / Editing and style transfer
Attention-GAN for Object Transfiguration in Wild Images
Xinyuan Chen, Chang Xu, Xiaokang Yang, Dacheng Tao
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 |
|---|---|---|---|
| split generator into attention and transformation networks (vs CycleGAN) | CycleGAN · MSCOCO horse to zebra | background PSNR: 18.1875 -> 22.2629 | Improved |
| split generator into attention and transformation networks (vs CycleGAN) | CycleGAN · MSCOCO horse to zebra | background SSIM: 0.6725 -> 0.9003 | Improved |
| split generator into attention and transformation networks (vs CycleGAN) | CycleGAN · MSCOCO zebra to horse | background PSNR: 18.1021 -> 21.5360 | Improved |
| split generator into attention and transformation networks (vs CycleGAN) | CycleGAN · MSCOCO zebra to horse | background SSIM: 0.7155 -> 0.8988 | Improved |
| add segmentation supervision to attention network (ablation) | Attention-GAN (unsupervised) · MSCOCO horse to zebra | background PSNR: 22.2629 -> 24.589 | Improved |
| add segmentation supervision to attention network (ablation) | Attention-GAN (unsupervised) · MSCOCO horse to zebra | background SSIM: 0.9003 -> 0.9482 | Improved |
| add segmentation supervision to attention network (ablation) | Attention-GAN (unsupervised) · MSCOCO zebra to horse | background PSNR: 21.5360 -> 23.9330 | Improved |
| add segmentation supervision to attention network (ablation) | Attention-GAN (unsupervised) · MSCOCO horse foreground mask | UoI (%): 28.1% -> 37.8% | Improved |
| add segmentation supervision to attention network (ablation) | Attention-GAN (unsupervised) · MSCOCO horse foreground mask | mAPr@0.5 (%): 20.3% -> 30.5% | Improved |
| add sparse loss lambda_sparse=1 (vs 0) (ablation) | Attention-GAN (unsupervised) · MSCOCO horse to zebra | background PSNR: 19.8621 -> 22.2629 | Improved |
| increase sparse loss weight from 1 to 5 (ablation) | Attention-GAN (unsupervised) · MSCOCO horse to zebra | background PSNR: 22.2629 -> 24.2173 | Mixed |
| increase sparse loss weight from 1 to 5 (ablation) | Attention-GAN (unsupervised) · MSCOCO horse to zebra | background SSIM: 0.9003 -> 0.9367 | Mixed |
Closest papers by meaning
- Unsupervised Attention-guided Image-to-Image Translation (NeurIPS 2018 · similarity 0.89)
- DA-GAN: Instance-Level Image Translation by Deep Attention Generative Adversarial Networks (CVPR 2018 · similarity 0.88)
- Paying More Attention to Attention: Improving the Performance of Convolutional Neural Networks via Attention Transfer (ICLR 2017 · similarity 0.87)
- U-GAT-IT: Unsupervised Generative Attentional Networks with Adaptive Layer-Instance Normalization for Image-to-Image Translation (ICLR 2020 · similarity 0.86)
- LookHere: Vision Transformers with Directed Attention Generalize and Extrapolate (NeurIPS 2024 · similarity 0.86)
- 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.
