Group Normalization
Explore this paper's citation graph
Summary
Group Normalization (GN) is presented as a simple alternative to BN that can outperform its BN-based counterparts for object detection and segmentation in COCO, and for video classification in Kinetics, showing that GN can effectively replace the powerful BN in a variety of tasks.
- Type
- preprint
- Published
- 2018-03-22
- Cited by
- 4,550
- References
- 75
- Access
- Open access
- OpenAlex
- https://openalex.org/W2795783309
- Semantic Scholar
- https://api.semanticscholar.org/CorpusID:4076251
Keywords
Normalization (sociology), Computer science, Segmentation, Pattern recognition (psychology), Computation
References
- Natural signal statistics and sensory gain control
- Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks
- Learning representations by back-propagating errors
- Learning Spatiotemporal Features with 3D Convolutional Networks
- Understanding the difficulty of training deep feedforward neural networks
- Modeling the Shape of the Scene: A Holistic Representation of the Spatial Envelope
- One weird trick for parallelizing convolutional neural networks
- Delving Deep into Rectifiers: Surpassing Human-Level Performance on ImageNet Classification
- Very Deep Convolutional Networks for Large-Scale Image Recognition
- Fully convolutional networks for semantic segmentation
- Normalization as a canonical neural computation
- Aggregating local descriptors into a compact image representation
- Long Short-Term Memory
- Going deeper with convolutions
- Object Detection Networks on Convolutional Feature Maps
- ImageNet Large Scale Visual Recognition Challenge
- Nonlinear image representation using divisive normalization
- Fisher Kernels on Visual Vocabularies for Image Categorization
- Recognizing linked events: Searching the space of feasible explanations
- Natural image statistics and neural representation.
Cited by
- Domain Adaptation Using Adversarial Learning for Autonomous Navigation
- Substitute Teacher Networks: Learning with Almost No Supervision
- Revisiting Small Batch Training for Deep Neural Networks
- Training Faster by Separating Modes of Variation in Batch-Normalized Models
- How Does Batch Normalization Help Optimization? (No, It Is Not About Internal Covariate Shift)
- End-to-End Parkinson Disease Diagnosis using Brain MR-Images by 3D-CNN
- Laplacian smoothing gradient descent
- Differentiable Learning-to-Normalize via Switchable Normalization
- 3-D RoI-Aware U-Net for Accurate and Efficient Colorectal Tumor Segmentation
- Large-Scale Visual Speech Recognition
- Non-local RoIs for Instance Segmentation
- Icing on the Cake: An Easy and Quick Post-Learnig Method You Can Try After Deep Learning
- Collapse of Deep and Narrow Neural Nets
- The Impact of Regularization on Convolutional Neural Networks
- Unsupervised Learning of Dense Optical Flow and Depth from Sparse Event Data
- Normalization in Training Deep Convolutional Neural Networks for 2D Bio-medical Semantic Segmentation
- CORnet: Modeling the Neural Mechanisms of Core Object Recognition
- Direct Training for Spiking Neural Networks: Faster, Larger, Better
- Mini-batch Serialization: CNN Training with Inter-layer Data Reuse
- Unsupervised Object Detection in Images from Maritime Environments