Network In Network
Explore this paper's citation graph
Summary
With enhanced local modeling via the micro network, the proposed deep network structure NIN is able to utilize global average pooling over feature maps in the classification layer, which is easier to interpret and less prone to overfitting than traditional fully connected layers.
- Type
- article
- Published
- 2013-12-16
- Cited by
- 6,777
- References
- 20
- Access
- Open access
- OpenAlex
- https://openalex.org/W2963911037
- Semantic Scholar
- https://api.semanticscholar.org/CorpusID:16636683
Keywords
MNIST database, Overfitting, Computer science, Artificial intelligence, Activation function
References
- Regularization of Neural Networks using DropConnect
- Knowledge Matters: Importance of Prior Information for Optimization
- Learnable Pooling Regions for Image Classification
- Improving neural networks by preventing co-adaptation of feature detectors
- Piecewise Linear Multilayer Perceptrons and Dropout
- PRINCIPLES OF NEURODYNAMICS. PERCEPTRONS AND THE THEORY OF BRAIN MECHANISMS
- Learning Hierarchical Features for Scene Labeling
- Discriminative Transfer Learning with Tree-based Priors
- ICA with Reconstruction Cost for Efficient Overcomplete Feature Learning
- Gradient-based learning applied to document recognition
- Human Face Detection in Visual Scenes
- Practical Bayesian Optimization of Machine Learning Algorithms
- ImageNet classification with deep convolutional neural networks
- Representation Learning: A Review and New Perspectives
- Improving Neural Networks with Dropout
- Reading Digits in Natural Images with Unsupervised Feature Learning
- Multi-digit Number Recognition from Street View Imagery using Deep Convolutional Neural Networks
- Stochastic Pooling for Regularization of Deep Convolutional Neural Networks
- Visualizing and Understanding Convolutional Networks
- Learning Multiple Layers of Features from Tiny Images
Cited by
- Generalizing Pooling Functions in CNNs: Mixed, Gated, and Tree
- Skin Lesion Analysis towards Melanoma Detection Using Deep Learning Network
- Features Combined From Hundreds of Midlayers: Hierarchical Networks With Subnetwork Nodes
- Spatially-sparse convolutional neural networks
- Fully Convolutional Neural Networks for Crowd Segmentation
- DropSample: A New Training Method to Enhance Deep Convolutional Neural Networks for Large-Scale Unconstrained Handwritten Chinese Character Recognition
- APAC: Augmented PAttern Classification with Neural Networks
- Deep Representation Learning with Target Coding
- Convolutional Channel Features
- How Well Can a CNN Marginalize Simple Nuisances It is Designed for?
- Bayesian Convolutional Neural Networks with Bernoulli Approximate Variational Inference
- Supervised Learning of Semantics-Preserving Hashing via Deep Neural Networks for Large-Scale Image Search
- Training Very Deep Networks
- Towards dropout training for convolutional neural networks
- You Only Look Once: Unified, Real-Time Object Detection
- Pedestrian detection with a Large-Field-Of-View deep network
- PerforatedCNNs: Acceleration through Elimination of Redundant Convolutions
- A discriminative cascade CNN model for offline handwritten digit recognition
- Deep Convolution Neural Networks in Computer Vision: a Review
- HCP: A Flexible CNN Framework for Multi-Label Image Classification
Related papers
- Learning Multiple Layers of Features from Tiny Images
- Deep Residual Learning for Image Recognition
- Rethinking the Inception Architecture for Computer Vision
- ImageNet classification with deep convolutional neural networks
- Caffe: Convolutional Architecture for Fast Feature Embedding
- ImageNet Large Scale Visual Recognition Challenge