Research map / Image generation / Image restoration
Grid Anchor Based Image Cropping: A New Benchmark and An Efficient Model
Hui Zeng, Lida Li, Zisheng Cao, Lei Zhang
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 multi-scale features (ablation) | GAIC VGG16 · GAICD (200 test images) | SRCC: 0.764 -> 0.777 | Improved |
| add multi-scale features (ablation) | GAIC MobileNetV2 · GAICD (200 test images) | SRCC: 0.775 -> 0.783 | Improved |
| add multi-scale features (ablation) | GAIC ShuffleNetV2 · GAICD (200 test images) | SRCC: 0.763 -> 0.774 | Improved |
| add data augmentation (ablation) | GAIC MobileNetV2 · GAICD (200 test images) | SRCC: 0.760 -> 0.775 | Improved |
| add data augmentation (ablation) | GAIC VGG16 · GAICD (200 test images) | SRCC: 0.752 -> 0.764 | Improved |
| replace VGG16 with MobileNetV2 backbone | GAIC multi-scale · GAICD (200 test images) | SRCC: 0.777 -> 0.783 | Improved |
| add RoD modeling to RoI-only (ablation) | GAIC MobileNetV2 · GAICD (200 test images) | SRCC: 0.770 -> 0.783 | Improved |
| add RoI modeling to RoD-only (ablation) | GAIC MobileNetV2 · GAICD (200 test images) | SRCC: 0.672 -> 0.770 | Improved |
| add RoD modeling to RoI-only (ablation) | GAIC ShuffleNetV2 · GAICD (200 test images) | SRCC: 0.764 -> 0.774 | Improved |
| use one 9x9 FC layer instead of three 3x3 conv layers (ablation) | GAIC ShuffleNetV2 · GAICD (200 test images) | SRCC: 0.765 -> 0.774 | Improved |
| use one 9x9 FC layer instead of two 5x5 conv layers (ablation) | GAIC ShuffleNetV2 · GAICD (200 test images) | SRCC: 0.769 -> 0.774 | Improved |
| increase channel dimension from 2 to 16 (ablation) | GAIC ShuffleNetV2 · GAICD (200 test images) | SRCC: 0.752 -> 0.774 | Improved |
Closest papers by meaning
- Reliable and Efficient Image Cropping: A Grid Anchor Based Approach (CVPR 2019 · similarity 0.98)
- Rethinking Image Cropping: Exploring Diverse Compositions From Global Views (CVPR 2022 · similarity 0.88)
- Spatial-Semantic Collaborative Cropping for User Generated Content (AAAI 2024 · similarity 0.87)
- Cropper: Vision-Language Model for Image Cropping through In-Context Learning (CVPR 2025 · similarity 0.86)
- Learning Subject-Aware Cropping by Outpainting Professional Photos (AAAI 2024 · similarity 0.86)
- Venus: Benchmarking and Empowering Multimodal Large Language Models for Aesthetic Guidance and Cropping (CVPR 2026 · similarity 0.83)
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.
