Proxy-Normalizing Activations to Match Batch Normalization while Removing Batch Dependence
arXiv:2106.03743
Abstract
We investigate the reasons for the performance degradation incurred with batch-independent normalization. We find that the prototypical techniques of layer normalization and instance normalization both induce the appearance of failure modes in the neural network's pre-activations: (i) layer normalization induces a collapse towards channel-wise constant functions; (ii) instance normalization induces a lack of variability in instance statistics, symptomatic of an alteration of the expressivity. To alleviate failure mode (i) without aggravating failure mode (ii), we introduce the technique "Proxy Normalization" that normalizes post-activations using a proxy distribution. When combined with layer normalization or group normalization, this batch-independent normalization emulates batch normalization's behavior and consistently matches or exceeds its performance.
arXiv commentNeurIPS 2021 camera-ready
References in corpus (31)
- Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift
- Instance Normalization: The Missing Ingredient for Fast Stylization
- Accurate, Large Minibatch SGD: Training ImageNet in 1 Hour
- A Learned Representation For Artistic Style
- Texture Networks: Feed-forward Synthesis of Textures and Stylized Images
- Layer Normalization
- Revisiting Small Batch Training for Deep Neural Networks
- Modulating early visual processing by language
- High-Performance Large-Scale Image Recognition Without Normalization
- Bayesian Uncertainty Estimation for Batch Normalized Deep Networks
- Deep ReLU Networks Have Surprisingly Few Activation Patterns
- An Investigation into Neural Net Optimization via Hessian Eigenvalue Density
- Norm matters: efficient and accurate normalization schemes in deep networks
- Theoretical Analysis of Auto Rate-Tuning by Batch Normalization
- Comparison of Batch Normalization and Weight Normalization Algorithms for the Large-scale Image Classification
- Complexity of Linear Regions in Deep Networks
- Batch Normalization is a Cause of Adversarial Vulnerability
- Towards Understanding Regularization in Batch Normalization
- Normalizing the Normalizers: Comparing and Extending Network Normalization Schemes
- Four Things Everyone Should Know to Improve Batch Normalization
- Online Normalization for Training Neural Networks
- Towards Stabilizing Batch Statistics in Backward Propagation of Batch Normalization
- Rethinking "Batch" in BatchNorm
- Batch Normalization Provably Avoids Rank Collapse for Randomly Initialised Deep Networks
- Mode Normalization
- Normalization of Neural Networks using Analytic Variance Propagation
- Characterizing Well-Behaved vs. Pathological Deep Neural Networks
- Making EfficientNet More Efficient: Exploring Batch-Independent Normalization, Group Convolutions and Reduced Resolution Training
- Mean Shift Rejection: Training Deep Neural Networks Without Minibatch Statistics or Normalization
- Accelerating Training of Deep Neural Networks with a Standardization Loss
- Is Batch Norm unique? An empirical investigation and prescription to emulate the best properties of common normalizers without batch dependence