Research map / Image generation / Image restoration
Deep Rectangling for Image Stitching: A Learning Baseline
Lang Nie, Chunyu Lin, Kang Liao, Shuaicheng Liu, Yao Zhao
Check 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 |
|---|---|---|---|
| Replace traditional two-stage rectangling with the learned one-stage mesh prediction | He et al. rectangling · DIR-D test | FID: 38.19 -> 21.77 | Improved |
| Replace traditional two-stage rectangling with the learned one-stage mesh prediction | He et al. rectangling · DIR-D test | SSIM: 0.3775 -> 0.7141 | Improved |
| Replace traditional two-stage rectangling with the learned one-stage mesh prediction | He et al. rectangling · DIR-D test | PSNR: 14.70 -> 21.28 | Improved |
| Add perceptual content loss to appearance and boundary losses (ablation) | Deep Rectangling · DIR-D test | FID: 115.37 -> 24.57 | Improved |
| Add mesh-shape loss to appearance, perception, and boundary losses (ablation) | Deep Rectangling · DIR-D test | FID: 24.57 -> 22.43 | Improved |
| Use a 4×3 mesh instead of an 8×6 mesh (ablation) | Deep Rectangling · DIR-D test | FID: 22.43 -> 24.15 | Hurt |
| Use a 16×12 mesh instead of an 8×6 mesh (ablation) | Deep Rectangling · DIR-D test | FID: 22.43 -> 22.32 | No clear effect |
| Add residual mesh refinement after primary regression (ablation) | Deep Rectangling · DIR-D test | SSIM: 0.6902 -> 0.7141 | Improved |
| Add residual mesh refinement after primary regression (ablation) | Deep Rectangling · DIR-D test | FID: 22.35 -> 21.77 | Improved |
| Add mesh-shape and content-preserving constraints (ablation) | Deep Rectangling · DIR-D test | PSNR: 14.43 -> 20.92 | Improved |
Closest papers by meaning
- Reconstructing the Image Stitching Pipeline: Integrating Fusion and Rectangling into a Unified Inpainting Model (NeurIPS 2024 · similarity 0.89)
- RecDiffusion: Rectangling for Image Stitching with Diffusion Models (CVPR 2024 · similarity 0.89)
- Depth-Supervised Fusion Network for Seamless-Free Image Stitching (NeurIPS 2025 · similarity 0.88)
- Towards Robust Image Stitching: An Adaptive Resistance Learning against Compatible Attacks (AAAI 2024 · similarity 0.86)
- Object-level Geometric Structure Preserving for Natural Image Stitching (AAAI 2025 · similarity 0.85)
- Decompositional Neural Scene Reconstruction with Generative Diffusion Prior (CVPR 2025 · similarity 0.84)
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.
