Batch-normalized Recurrent Highway Networks

Abstract

Gradient control plays an important role in feed-forward networks applied to various computer vision tasks. Previous work has shown that Recurrent Highway Networks minimize the problem of vanishing or exploding gradients. They achieve this by setting the eigenvalues of the temporal Jacobian to 1 across the time steps. In this work, batch normalized recurrent highway networks are proposed to control the gradient flow in an improved way for network convergence. Specifically, the introduced model can be formed by batch normalizing the inputs at each recurrence loop. The proposed model is tested on an image captioning task using MSCOCO dataset. Experimental results indicate that the batch normalized recurrent highway networks converge faster and performs better compared with the traditional LSTM and RHN based models.

arXiv comment5 pages, 3 figures, Published in 2017 IEEE International Conference on Image Processing (ICIP)

References in corpus (11)

• Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift
• LSTM: A Search Space Odyssey
• Microsoft COCO Captions: Data Collection and Evaluation Server
• Character-Aware Neural Language Models
• Show and Tell: Lessons learned from the 2015 MSCOCO Image Captioning Challenge
• Weight Normalization: A Simple Reparameterization to Accelerate Training of Deep Neural Networks
• Highway Networks
• Deep Networks with Stochastic Depth
• How to Construct Deep Recurrent Neural Networks
• Recurrent Batch Normalization
• Convolutional Residual Memory Networks