Improving Model Accuracy for Imbalanced Image Classification Tasks by Adding a Final Batch Normalization Layer: An Empirical Study
Explore this paper's citation graph
Summary
It is demonstrated that utilizing an additional BN layer before the output layer in modern CNN architectures has a considerable impact in terms of minimizing the training time and testing error for minority classes in highly imbalanced data sets.
- Type
- preprint
- Published
- 2020-11-12
- Cited by
- 7
- References
- 36
- Access
- Open access
- OpenAlex
- https://openalex.org/W3103237175
- Semantic Scholar
- https://api.semanticscholar.org/CorpusID:226307041
Keywords
Normalization (sociology), Computer science, Artificial intelligence, Skewness, Machine learning
References
- Very Deep Convolutional Networks for Large-Scale Image Recognition
- Evolution and Management of the Irish Potato Famine Pathogen Phytophthora Infestans in Canada and the United States
- Recent advances in sensing plant diseases for precision crop protection
- VERIFICATION OF FORECASTS EXPRESSED IN TERMS OF PROBABILITY
- A Simple Weight Decay Can Improve Generalization
- Human-level concept learning through probabilistic program induction
- Deep Residual Learning for Image Recognition
- An open access repository of images on plant health to enable the development of mobile disease diagnostics through machine learning and crowdsourcing
- Skin lesion analysis toward melanoma detection: A challenge at the 2017 International symposium on biomedical imaging (ISBI), hosted by the international skin imaging collaboration (ISIC)
- Using Deep Learning for Image-Based Plant Disease Detection
- Instance Normalization: The Missing Ingredient for Fast Stylization
- On Calibration of Modern Neural Networks
- Group Normalization
- How Does Batch Normalization Help Optimization? (No, It Is Not About Internal Covariate Shift)
- Batch Normalization: Is Learning An Adaptive Gain and Bias Necessary?
- Evaluating Late Blight Severity in Potato Crops Using Unmanned Aerial Vehicles and Machine Learning Algorithms
- Adaptive Estimators Show Information Compression in Deep Neural Networks
- Robust Anomaly Detection in Images using Adversarial Autoencoders
- Deep Learning on Small Datasets without Pre-Training using Cosine Loss
- Road Crack Detection Using Deep Convolutional Neural Network and Adaptive Thresholding
Cited by
- Quality Judgment of 3D Face Point Cloud Based on Feature Fusion
- A Coupled Compression Generation Network for Remote-Sensing Images at Extremely Low Bitrates
- A Survey of Malware Detection Using Deep Learning
- Exploring Batch Normalization’s Impact on Dense Layers of Multiclass and Multilabel Classifiers
- A Hybrid CNN-SVM Algorithm for Detecting Manufacturing Defects
- The Unreasonable Effectiveness of the Final Batch Normalization Layer
- Exploring Batch Normalization's Impact on Dense Layers of Multi-Class and Multi-Label Classifiers
Related papers
- A New Imbalanced Learning and Dictions Tree Method for Breast Cancer Diagnosis
- Learning From Imbalanced Data: Rank Metrics and Extra Tasks
- Imbalanced Image Classification with Complement Cross Entropy
- Enhancing techniques for learning decision trees from imbalanced data
- Cost Sensitive and Preprocessing for Classification with Imbalanced Data-sets: Similar Behaviour and Potential Hybridizations
- Minority Class Oversampling for Tabular Data with Deep Generative Models