Research map / Image generation / Editing and style transfer
IICNet: A Generic Framework for Reversible Image Conversion
Ka Leong Cheng, Yueqi Xie, Qifeng Chen
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 |
|---|---|---|---|
| add the relation module (ablation) | IICNet · video embedding | embedding PSNR: 38.752 -> 38.900 | Improved |
| add the relation module (ablation) | IICNet · video embedding | restored PSNR: 41.159 -> 41.729 | Improved |
| add the relation module (ablation) | IICNet · mononizing binocular images | restored PSNR: 39.667 -> 41.649 | Improved |
| add the relation module (ablation) | IICNet · hiding images in an image | restored PSNR: 36.698 -> 37.935 | Improved |
| add the frequency loss (ablation) | IICNet · video embedding | embedding PSNR: 32.914 -> 38.900 | Improved |
| add the frequency loss (ablation) | IICNet · video embedding | restored PSNR: 42.353 -> 41.729 | Hurt |
| add the frequency loss (ablation) | IICNet · hiding images in an image | embedding PSNR: 28.780 -> 35.641 | Improved |
| add the frequency loss (ablation) | IICNet · mononizing binocular images | embedding PSNR: 31.469 -> 37.540 | Improved |
| use IICNet instead of an auto-encoder (AE) baseline (ablation) | IICNet · hiding images in an image | restored PSNR: 31.721 -> 37.935 | Improved |
| use IICNet instead of the INN-with-auxiliary-maps baseline (ablation) | IICNet · video embedding | embedding PSNR: 34.029 -> 38.900 | Improved |
| use IICNet instead of Zhu et al. at time step 5 | IICNet · DAVIS 2017 video embedding | embedding PSNR: 24.246 -> 36.739 | Improved |
| increase embedding range from 5 to 9 images | IICNet · DAVIS 2017 video embedding | restored PSNR: 41.729 -> 36.698 | Hurt |
Later papers that compared with this method
2 later accepted papers report IICNet as a baseline in their results tables; 2 of them report a better number on most of the results they share with it.
- Tackling Ill-Posedness of Reversible Image Conversion With Well-Posed Invertible Network (TPAMI 2026 · better on 8 of 8 shared results)
- Robust Invertible Image Steganography (CVPR 2022 · better on 2 of 2 shared results)
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Tackling Ill-Posedness of Reversible Image Conversion With Well-Posed Invertible Network (TPAMI 2026 · similarity 0.91)
- i-RevNet: Deep Invertible Networks (ICLR 2018 · similarity 0.86)
- I2C: Invertible Continuous Codec for High-Fidelity Variable-Rate Image Compression (TPAMI 2024 · similarity 0.85)
- MintNet: Building Invertible Neural Networks with Masked Convolutions (NeurIPS 2019 · similarity 0.84)
- Reversible Decoupling Network for Single Image Reflection Removal (CVPR 2025 · similarity 0.83)
- Dr2Net: Dynamic Reversible Dual-Residual Networks for Memory-Efficient Finetuning (CVPR 2024 · similarity 0.82)
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.
