Pre-rotation Only at Inference-Stage: A Way to Rotation Invariance of Convolutional Neural Networks.
Yue FanPeng ZhangJingqi HanDandan LiuJinsong TangGuoping ZhangPublished in: Int. J. Comput. Intell. Syst. (2024)
Keyphrases
- rotation invariance
- convolutional neural networks
- rotation invariant
- scale invariance
- scale and rotation invariant
- fourier transform
- zernike moments
- texture analysis
- shape context
- texture features
- moment invariants
- local binary pattern
- texture classification
- image representation
- image analysis
- frequency domain
- image reconstruction
- multiscale
- texture images
- higher order
- state space
- multiresolution
- feature vectors
- pattern recognition