Capsule Network Performance on Complex Data

Abstract

In recent years, convolutional neural networks (CNN) have played an important role in the field of deep learning. Variants of CNN's have proven to be very successful in classification tasks across different domains. However, there are two big drawbacks to CNN's: their failure to take into account of important spatial hierarchies between features, and their lack of rotational invariance. As long as certain key features of an object are present in the test data, CNN's classify the test data as the object, disregarding features' relative spatial orientation to each other. This causes false positives. The lack of rotational invariance in CNN's would cause the network to incorrectly assign the object another label, causing false negatives. To address this concern, Hinton et al. propose a novel type of neural network using the concept of capsules in a recent paper. With the use of dynamic routing and reconstruction regularization, the capsule network model would be both rotation invariant and spatially aware. The capsule network has shown its potential by achieving a state-of-the-art result of 0.25% test error on MNIST without data augmentation such as rotation and scaling, better than the previous baseline of 0.39%. To further test out the application of capsule networks on data with higher dimensionality, we attempt to find the best set of configurations that yield the optimal test error on CIFAR10 dataset.

References in corpus (1)

• Dynamic Routing Between Capsules

Cited by in corpus (32)

• Efficient-CapsNet: Capsule Network with Self-Attention Routing
• Use of a Capsule Network to Detect Fake Images and Videos
• Investigating Capsule Networks with Dynamic Routing for Text Classification
• Recent advances for quantum classifiers
• CapsNet comparative performance evaluation for image classification
• Morphological classification of radio galaxies: Capsule Networks versus Convolutional Neural Networks
• Novel Deep Learning Model for Traffic Sign Detection Using Capsule Networks
• Novel Dual-Channel Long Short-Term Memory Compressed Capsule Networks for Emotion Recognition
• Dense and Diverse Capsule Networks: Making the Capsules Learn Better
• HitNet: a neural network with capsules embedded in a Hit-or-Miss layer, extended with hybrid data augmentation and ghost capsules
• Few Shot Speaker Recognition using Deep Neural Networks
• Generalized Capsule Networks with Trainable Routing Procedure
• Capsule Networks Need an Improved Routing Algorithm
• Self-Attention Capsule Networks for Object Classification
• HSI-CNN: A Novel Convolution Neural Network for Hyperspectral Image
• DESCNet: Developing Efficient Scratchpad Memories for Capsule Network Hardware
• Dynamic Layer Aggregation for Neural Machine Translation with Routing-by-Agreement
• On the Vulnerability of Capsule Networks to Adversarial Attacks
• Information Aggregation for Multi-Head Attention with Routing-by-Agreement
• Multi-labeled Relation Extraction with Attentive Capsule Network
• Examining the Benefits of Capsule Neural Networks
• Trustable and Automated Machine Learning Running with Blockchain and Its Applications
• A Context-aware Capsule Network for Multi-label Classification
• An Effective Hit-or-Miss Layer Favoring Feature Interpretation as Learned Prototypes Deformations
• Using dynamic routing to extract intermediate features for developing scalable capsule networks
• Wasserstein Routed Capsule Networks
• TimeCaps: Capturing Time Series Data With Capsule Networks
• Capsule Network Performance with Autonomous Navigation
• Enabling Highly Efficient Capsule Networks Processing Through A PIM-Based Architecture Design
• Graph Routing between Capsules
• P-CapsNets: a General Form of Convolutional Neural Networks
• Examining the causal structures of deep neural networks using information theory