Research map / Detection and segmentation / CNN features
RRL: Regional Rotate Layer in Convolutional Neural Networks
Zongbo Hao, Tao Zhang, Mingwang Chen, Zou Kaixu
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 RRL to LeNet-5 | LeNet-5 · CIFAR10-rot test, trained on upright CIFAR-10 | accuracy (%): 33.2 -> 71.3 | Improved |
| Add RRL to LeNet-5 | LeNet-5 · CIFAR10-rot+ test, trained on upright CIFAR-10 | accuracy (%): 18.2 -> 52.8 | Improved |
| Train LeNet-5+RRL on rotated data instead of upright data | LeNet-5+RRL · CIFAR10-rot test | accuracy (%): 71.3 -> 70.9 | No clear effect |
| Train LeNet-5+RRL on rotated data instead of upright data | LeNet-5+RRL · CIFAR10-rot+ test | accuracy (%): 52.8 -> 49.1 | Hurt |
| Add RRL to ResNet-18 | ResNet-18 · CIFAR10-rot test, trained on upright CIFAR-10 | accuracy (%): 46.5 -> 75.0 | Improved |
| Add RRL to ResNet-18 | ResNet-18 · CIFAR10-rot+ test, trained on upright CIFAR-10 | accuracy (%): 38.7 -> 65.3 | Improved |
| Train ResNet-18+RRL on rotated data instead of upright data | ResNet-18+RRL · CIFAR10-rot test | accuracy (%): 75.0 -> 77.9 | No clear effect |
| Train ResNet-18+RRL on rotated data instead of upright data | ResNet-18+RRL · CIFAR10-rot+ test | accuracy (%): 65.3 -> 63.1 | No clear effect |
| Add RRL to ResNet-44 | ResNet-44 · Plankton real test set | multi-class score: 3.67862 -> 2.18777 | Improved |
| Add RRL to MobileNet-tiny-yolov3 | MobileNet-tiny-yolov3 · Pascal VOC rotated-image test | mAP: 43.76% -> 61.78% | Improved |
Closest papers by meaning
- Rotation Equivariant Vector Field Networks (ICCV 2017 · similarity 0.90)
- Harmonic Networks: Deep Translation and Rotation Equivariance (CVPR 2017 · similarity 0.87)
- Achieving Rotational Invariance with Bessel-Convolutional Neural Networks (NeurIPS 2021 · similarity 0.87)
- Oriented Response Networks (CVPR 2017 · similarity 0.87)
- Exploiting Cyclic Symmetry in Convolutional Neural Networks (ICML 2016 · similarity 0.86)
- RotDCF: Decomposition of Convolutional Filters for Rotation-Equivariant Deep Networks (ICLR 2019 · 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.
