Scale Calibrated Training: Improving Generalization of Deep Networks via Scale-Specific Normalization
arXiv:1909.00182
Abstract
Standard convolutional neural networks(CNNs) require consistent image resolutions in both training and testing phase. However, in practice, testing with smaller image sizes is necessary for fast inference. We show that trivially evaluating low-resolution images on networks trained with high-resolution images results in a catastrophic accuracy drop in standard CNN architectures. We propose a novel training regime called Scale calibrated Training(SCT) which allows networks to learn from various scales of input simultaneously. By taking advantages of SCT, single network can provide decent accuracy at test time in response to multiple test scales. In our analysis, we surprisingly find that vanilla batch normalization can lead to sub-optimal performance in SCT. Therefore, a novel normalization scheme called Scale-Specific Batch Normalization is equipped to SCT in replacement of batch normalization. Experiment results show that SCT improves accuracy of single Resnet-50 on ImageNet by 1.7% and 11.5% accuracy when testing on image sizes of 224 and 128 respectively.
arXiv commentTechnical Report
References in corpus (18)
- Very Deep Convolutional Networks for Large-Scale Image Recognition
- Distilling the Knowledge in a Neural Network
- MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications
- EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks
- mixup: Beyond Empirical Risk Minimization
- Instance Normalization: The Missing Ingredient for Fast Stylization
- Improved Regularization of Convolutional Neural Networks with Cutout
- Deep Speech 2: End-to-End Speech Recognition in English and Mandarin
- Progressive Growing of GANs for Improved Quality, Stability, and Variation
- The Cityscapes Dataset for Semantic Urban Scene Understanding
- PACT: Parameterized Clipping Activation for Quantized Neural Networks
- Towards Dropout Training for Convolutional Neural Networks
- Modulating early visual processing by language
- Slimmable Neural Networks
- Differentiable Learning-to-Normalize via Switchable Normalization
- Fixup Initialization: Residual Learning Without Normalization
- Fixing the train-test resolution discrepancy
- SSN: Learning Sparse Switchable Normalization via SparsestMax