MEC: Memory-efficient Convolution for Deep Neural Network

Abstract

Convolution is a critical component in modern deep neural networks, thus several algorithms for convolution have been developed. Direct convolution is simple but suffers from poor performance. As an alternative, multiple indirect methods have been proposed including im2col-based convolution, FFT-based convolution, or Winograd-based algorithm. However, all these indirect methods have high memory-overhead, which creates performance degradation and offers a poor trade-off between performance and memory consumption. In this work, we propose a memory-efficient convolution or MEC with compact lowering, which reduces memory-overhead substantially and accelerates convolution process. MEC lowers the input matrix in a simple yet efficient/compact way (i.e., much less memory-overhead), and then executes multiple small matrix multiplications in parallel to get convolution completed. Additionally, the reduced memory footprint improves memory sub-system efficiency, improving performance. Our experimental results show that MEC reduces memory consumption significantly with good speedup on both mobile and server platforms, compared with other indirect convolution algorithms.

ICML2017

References in corpus (11)

• Deep Residual Learning for Image Recognition
• Deep Compression: Compressing Deep Neural Networks with Pruning, Trained Quantization and Huffman Coding
• OverFeat: Integrated Recognition, Localization and Detection using Convolutional Networks
• Compressing Deep Convolutional Networks using Vector Quantization
• Speeding up Convolutional Neural Networks with Low Rank Expansions
• Fast Training of Convolutional Networks through FFTs
• Fast Convolutional Nets With fbfft: A GPU Performance Evaluation
• Memory Bounded Deep Convolutional Networks
• Compression of Deep Convolutional Neural Networks for Fast and Low Power Mobile Applications
• Fast Algorithms for Convolutional Neural Networks
• Caffe con Troll: Shallow Ideas to Speed Up Deep Learning

Cited by in corpus (12)

• Deep Learning in Mobile and Wireless Networking: A Survey
• RepMLP: Re-parameterizing Convolutions into Fully-connected Layers for Image Recognition
• Crossbar-aware neural network pruning
• Photonic Convolution Neural Network Based on Interleaved Time-Wavelength Modulation
• RNNPool: Efficient Non-linear Pooling for RAM Constrained Inference
• Towards a learning-based performance modeling for accelerating Deep Neural Networks
• Deep Learning Training on the Edge with Low-Precision Posits
• Im2win: Memory Efficient Convolution On SIMD Architectures
• High Performance Depthwise and Pointwise Convolutions on Mobile Devices
• TASO: Time and Space Optimization for Memory-Constrained DNN Inference
• Low-memory convolutional neural networks through incremental depth-first processing
• Modulating Regularization Frequency for Efficient Compression-Aware Model Training