ReNet: A Recurrent Neural Network Based Alternative to Convolutional Networks

Abstract

In this paper, we propose a deep neural network architecture for object recognition based on recurrent neural networks. The proposed network, called ReNet, replaces the ubiquitous convolution+pooling layer of the deep convolutional neural network with four recurrent neural networks that sweep horizontally and vertically in both directions across the image. We evaluate the proposed ReNet on three widely-used benchmark datasets; MNIST, CIFAR-10 and SVHN. The result suggests that ReNet is a viable alternative to the deep convolutional neural network, and that further investigation is needed.

References in corpus (17)

• Adam: A Method for Stochastic Optimization
• Sequence to Sequence Learning with Neural Networks
• Show, Attend and Tell: Neural Image Caption Generation with Visual Attention
• Improving neural networks by preventing co-adaptation of feature detectors
• Practical Bayesian Optimization of Machine Learning Algorithms
• On the difficulty of training Recurrent Neural Networks
• Striving for Simplicity: The All Convolutional Net
• Going Deeper with Convolutions
• Theano: new features and speed improvements
• One weird trick for parallelizing convolutional neural networks
• Stochastic Pooling for Regularization of Deep Convolutional Neural Networks
• Gated Feedback Recurrent Neural Networks
• End-to-end Continuous Speech Recognition using Attention-based Recurrent NN: First Results
• Fractional Max-Pooling
• Spatially-sparse convolutional neural networks
• Describing Videos by Exploiting Temporal Structure
• Convolutional Kernel Networks

Cited by in corpus (50)

• A Review on Deep Learning Techniques Applied to Semantic Segmentation
• Spatial-Temporal Recurrent Neural Network for Emotion Recognition
• Recent Advances in Recurrent Neural Networks
• Grid Long Short-Term Memory
• Self-Taught Convolutional Neural Networks for Short Text Clustering
• What-and-Where to Match: Deep Spatially Multiplicative Integration Networks for Person Re-identification
• Learning Contextual Dependencies with Convolutional Hierarchical Recurrent Neural Networks
• MelNet: A Generative Model for Audio in the Frequency Domain
• How Much Position Information Do Convolutional Neural Networks Encode?
• SAC-Net: Spatial Attenuation Context for Salient Object Detection
• Inside-Outside Net: Detecting Objects in Context with Skip Pooling and Recurrent Neural Networks
• TabularNet: A Neural Network Architecture for Understanding Semantic Structures of Tabular Data
• An End-to-End Breast Tumour Classification Model Using Context-Based Patch Modelling- A BiLSTM Approach for Image Classification
• High-Resolution Shape Completion Using Deep Neural Networks for Global Structure and Local Geometry Inference
• A Comprehensive Review of Modern Object Segmentation Approaches
• Combining the Best of Convolutional Layers and Recurrent Layers: A Hybrid Network for Semantic Segmentation
• Position, Padding and Predictions: A Deeper Look at Position Information in CNNs
• Reading Scene Text in Deep Convolutional Sequences
• Semantic Image Segmentation with Task-Specific Edge Detection Using CNNs and a Discriminatively Trained Domain Transform
• Pedestrian Attribute Recognition: A Survey
• RNNPool: Efficient Non-linear Pooling for RAM Constrained Inference
• Learning Affinity via Spatial Propagation Networks
• A Deep Spatial Contextual Long-term Recurrent Convolutional Network for Saliency Detection
• Convolutional RNN: an Enhanced Model for Extracting Features from Sequential Data
• Structure-Aware Network for Lane Marker Extraction with Dynamic Vision Sensor
• Richer and Deeper Supervision Network for Salient Object Detection
• Image-based localization using LSTMs for structured feature correlation
• How Can CNNs Use Image Position for Segmentation?
• JuncNet: A Deep Neural Network for Road Junction Disambiguation for Autonomous Vehicles
• A Survey on Deep Learning Methods for Semantic Image Segmentation in Real-Time
• A Dilated Inception Network for Visual Saliency Prediction
• DartsReNet: Exploring new RNN cells in ReNet architectures
• A Deep Neural Network Surrogate Modeling Benchmark for Temperature Field Prediction of Heat Source Layout
• A Deep Neuro-Fuzzy Network for Image Classification
• Spatial Dependency Networks: Neural Layers for Improved Generative Image Modeling
• Face Parsing via Recurrent Propagation
• CIFAR-10: KNN-based Ensemble of Classifiers
• Scene Labeling using Gated Recurrent Units with Explicit Long Range Conditioning
• ContinuityLearner: Geometric Continuity Feature Learning for Lane Segmentation
• On the Privacy Risks of Deploying Recurrent Neural Networks in Machine Learning Models
• Multilevel Context Representation for Improving Object Recognition
• Consensus Feature Network for Scene Parsing
• User Constrained Thumbnail Generation using Adaptive Convolutions
• Supervised Deep Sparse Coding Networks
• Learning from Counting: Leveraging Temporal Classification for Weakly Supervised Object Localization and Detection
• Text Classification based on Multiple Block Convolutional Highways
• Deep Semantics-Aware Photo Adjustment
• LSTM-CF: Unifying Context Modeling and Fusion with LSTMs for RGB-D Scene Labeling
• Inspect Transfer Learning Architecture with Dilated Convolution
• SaLite : A light-weight model for salient object detection