Research map / Architectures and efficiency / CNN design
Efficient Classification of Very Large Images With Tiny Objects
Fanjie Kong, Ricardo Henao
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 one-stage ATS sampling with two-stage hierarchical attention sampling | Zoom-In Network · Colon Cancer | accuracy: 90.7 ± 1.4 -> 95.0 ± 2.6 | Improved |
| Replace one-stage ATS sampling with two-stage hierarchical attention sampling | Zoom-In Network · NeedleCamelyon | accuracy: 72.5 -> 76.0 | Improved |
| Use ResNet16 instead of LeNet as the feature-extractor backbone | Zoom-In Network · NeedleCamelyon | accuracy: 76.0 -> 78.1 | Improved |
| Use ResNet16 instead of LeNet as the feature-extractor backbone | Zoom-In Network · Traffic Signs Recognition | accuracy: 91.2 -> 92.6 | Improved |
| Use ResNet16 instead of LeNet as the feature-extractor backbone | Zoom-In Network · Functional Map of the World | accuracy: 72.9 -> 74.3 | Improved |
| Use ResNet16 instead of LeNet as the feature-extractor backbone | Zoom-In Network · Camelyon16 | accuracy: 81.3 -> 82.6 | Improved |
| Add pixel-level annotations to training | Zoom-In Network · Camelyon16 | accuracy: 81.3 -> 88.2 | Improved |
| Add contrastive learning (ablation) | Zoom-In Network · Colon Cancer | accuracy: 94.0 -> 95.0 | Improved |
| Add contrastive learning (ablation) | Zoom-In Network · NeedleCamelyon | accuracy: 74.3 -> 76.0 | Improved |
| Add contrastive learning (ablation) | Zoom-In Network · Camelyon16 | accuracy: 80.6 -> 81.3 | Improved |
| Set entropy regularization λ to 0 instead of 1e-5 (ablation) | Zoom-In Network · Colon Cancer | accuracy: 95.0 -> 94.0 | Hurt |
| Set entropy regularization λ to 1.0 instead of 1e-5 (ablation) | Zoom-In Network · Colon Cancer | accuracy: 95.0 -> 95.0 | No clear effect |
Closest papers by meaning
- Streaming Convolutional Neural Networks for End-to-End Learning With Multi-Megapixel Images (TPAMI 2020 · similarity 0.89)
- A Dataset Generation Framework for Evaluating Megapixel Image Classifiers & Their Explanations (ECCV 2022 · similarity 0.86)
- Neural Image Compression for Gigapixel Histopathology Image Analysis (TPAMI 2019 · similarity 0.86)
- Learn to Pay Attention (ICLR 2018 · similarity 0.85)
- Multi-Scale Dense Networks for Resource Efficient Image Classification (ICLR 2018 · similarity 0.85)
- Sequential Attention-based Sampling for Histopathological Analysis (NeurIPS 2025 · 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.
