Research map / Image generation / Image restoration
Classification-Driven Dynamic Image Enhancement
Vivek Sharma, Ali Diba, Davy Neven, Michael S. Brown, Luc Van Gool, Rainer Stiefelhagen
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 dynamically generated enhancement filters | AlexNet · PASCAL-VOC2007 | mAP: 76.9 -> 83.06 | Improved |
| Add dynamically generated enhancement filters | VGG-16 · PASCAL-VOC2007 | mAP: 89.3 -> 92.8 | Improved |
| Add dynamically generated enhancement filters | AlexNet · MIT-Indoor | accuracy: 56.79 -> 62.9 | Improved |
| Add dynamically generated enhancement filters | VGG-16 · MIT-Indoor | accuracy: 64.87 -> 68.67 | Improved |
| Add dynamically generated enhancement filters | AlexNet · DTD | mAP: 61.3 -> 67.81 | Improved |
| Add dynamically generated enhancement filters | VGG-VD · DTD | mAP: 67.0 -> 71.34 | Improved |
| Add dynamically generated enhancement filters | BN-Inception · CUB-200-2011 | accuracy: 82.3 -> 86.12 | Improved |
| Use 5x5 filters instead of 6x6 (ablation) | WLS dynamic enhancement on AlexNet · CUB-200-2011 | accuracy: 68.73 -> 65.9 | Hurt |
| Use 7x7 filters instead of 6x6 (ablation) | WLS dynamic enhancement on AlexNet · CUB-200-2011 | accuracy: 68.73 -> 66.84 | Hurt |
| Use MSE-based weights instead of averaging weights (ablation) | Stat-CNN · CUB-200-2011 | accuracy: 83.19 -> 83.74 | Improved |
| Use MSE-based weights instead of averaging weights (ablation) | Dyn-CNN · CUB-200-2011 | accuracy: 85.58 -> 86.12 | Improved |
| Train without regularization of weights (ablation) | Dyn-CNN · CUB-200-2011 | classification performance | Hurt |
Closest papers by meaning
- Training Compact CNNs for Image Classification Using Dynamic-Coded Filter Fusion (TPAMI 2023 · similarity 0.86)
- Dynamic Filter Networks (NeurIPS 2016 · similarity 0.86)
- CNN Filter DB: An Empirical Investigation of Trained Convolutional Filters (CVPR 2022 · similarity 0.86)
- An Empirical Evaluation of Current Convolutional Architectures' Ability to Manage Nuisance Location and Scale Variability (CVPR 2016 · similarity 0.86)
- Adaptive Dynamic Filtering Network for Image Denoising (AAAI 2023 · similarity 0.86)
- Learning Versatile Convolution Filters for Efficient Visual Recognition (TPAMI 2021 · similarity 0.86)
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.
