Rethinking Recurrent Neural Networks and Other Improvements for Image Classification
arXiv:2007.15161
Abstract
Over the long history of machine learning, which dates back several decades, recurrent neural networks (RNNs) have been used mainly for sequential data and time series and generally with 1D information. Even in some rare studies on 2D images, these networks are used merely to learn and generate data sequentially rather than for image recognition tasks. In this study, we propose integrating an RNN as an additional layer when designing image recognition models. We also develop end-to-end multimodel ensembles that produce expert predictions using several models. In addition, we extend the training strategy so that our model performs comparably to leading models and can even match the state-of-the-art models on several challenging datasets (e.g., SVHN (0.99), Cifar-100 (0.9027) and Cifar-10 (0.9852)). Moreover, our model sets a new record on the Surrey dataset (0.949). The source code of the methods provided in this article is available at https://github.com/leonlha/e2e-3m and http://nguyenhuuphong.me.
arXiv comment16 pages
References in corpus (25)
- Adam: A Method for Stochastic Optimization
- Very Deep Convolutional Networks for Large-Scale Image Recognition
- TensorFlow: Large-Scale Machine Learning on Heterogeneous Distributed Systems
- EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks
- End to End Learning for Self-Driving Cars
- Improved Regularization of Convolutional Neural Networks with Cutout
- A Critical Review of Recurrent Neural Networks for Sequence Learning
- Wide Residual Networks
- Pixel Recurrent Neural Networks
- Recurrent Neural Network for Text Classification with Multi-Task Learning
- DRAW: A Recurrent Neural Network For Image Generation
- Comparative Study of CNN and RNN for Natural Language Processing
- Recent Advances in Recurrent Neural Networks
- An End-to-End Spatio-Temporal Attention Model for Human Action Recognition from Skeleton Data
- Residual Networks of Residual Networks: Multilevel Residual Networks
- Fractional Max-Pooling
- Explaining How a Deep Neural Network Trained with End-to-End Learning Steers a Car
- ShakeDrop Regularization for Deep Residual Learning
- Population Based Augmentation: Efficient Learning of Augmentation Policy Schedules
- Style Augmentation: Data Augmentation via Style Randomization
- Improving Robustness Without Sacrificing Accuracy with Patch Gaussian Augmentation
- iCassava 2019 Fine-Grained Visual Categorization Challenge
- Competitive Multi-scale Convolution
- Mobile Machine Learning Hardware at ARM: A Systems-on-Chip (SoC) Perspective
- Drop-Activation: Implicit Parameter Reduction and Harmonic Regularization