DecomposeMe: Simplifying ConvNets for End-to-End Learning

Abstract

Deep learning and convolutional neural networks (ConvNets) have been successfully applied to most relevant tasks in the computer vision community. However, these networks are computationally demanding and not suitable for embedded devices where memory and time consumption are relevant. In this paper, we propose DecomposeMe, a simple but effective technique to learn features using 1D convolutions. The proposed architecture enables both simplicity and filter sharing leading to increased learning capacity. A comprehensive set of large-scale experiments on ImageNet and Places2 demonstrates the ability of our method to improve performance while significantly reducing the number of parameters required. Notably, on Places2, we obtain an improvement in relative top-1 classification accuracy of 7.7\% with an architecture that requires 92% fewer parameters compared to VGG-B. The proposed network is also demonstrated to generalize to other tasks by converting existing networks.

References in corpus (11)

• Very Deep Convolutional Networks for Large-Scale Image Recognition
• Distilling the Knowledge in a Neural Network
• Improving neural networks by preventing co-adaptation of feature detectors
• Practical Bayesian Optimization of Machine Learning Algorithms
• Deep Residual Learning for Image Recognition
• Exploiting Linear Structure Within Convolutional Networks for Efficient Evaluation
• One weird trick for parallelizing convolutional neural networks
• Stereo Matching by Training a Convolutional Neural Network to Compare Image Patches
• On the Number of Linear Regions of Deep Neural Networks
• Predicting Parameters in Deep Learning
• Speeding up Convolutional Neural Networks with Low Rank Expansions

Cited by in corpus (10)

• Context Based Emotion Recognition using EMOTIC Dataset
• Restricted Deformable Convolution based Road Scene Semantic Segmentation Using Surround View Cameras
• Compression-aware Training of Deep Networks
• The Devil is in the Decoder: Classification, Regression and GANs
• Binarized Convolutional Neural Networks with Separable Filters for Efficient Hardware Acceleration
• ESFNet: Efficient Network for Building Extraction from High-Resolution Aerial Images
• Binarized Convolutional Neural Networks for Efficient Inference on GPUs
• Elastic Neural Networks: A Scalable Framework for Embedded Computer Vision
• Fast Portrait Segmentation with Highly Light-weight Network
• Spatial Sampling Network for Fast Scene Understanding