Flattened Convolutional Neural Networks for Feedforward Acceleration

Abstract

We present flattened convolutional neural networks that are designed for fast feedforward execution. The redundancy of the parameters, especially weights of the convolutional filters in convolutional neural networks has been extensively studied and different heuristics have been proposed to construct a low rank basis of the filters after training. In this work, we train flattened networks that consist of consecutive sequence of one-dimensional filters across all directions in 3D space to obtain comparable performance as conventional convolutional networks. We tested flattened model on different datasets and found that the flattened layer can effectively substitute for the 3D filters without loss of accuracy. The flattened convolution pipelines provide around two times speed-up during feedforward pass compared to the baseline model due to the significant reduction of learning parameters. Furthermore, the proposed method does not require efforts in manual tuning or post processing once the model is trained.

International Conference on Learning Representations (ICLR) 2015

References in corpus (8)

• Very Deep Convolutional Networks for Large-Scale Image Recognition
• Caffe: Convolutional Architecture for Fast Feature Embedding
• Exploiting Linear Structure Within Convolutional Networks for Efficient Evaluation
• Going Deeper with Convolutions
• cuDNN: Efficient Primitives for Deep Learning
• Compressing Deep Convolutional Networks using Vector Quantization
• Speeding up Convolutional Neural Networks with Low Rank Expansions
• Speeding-up Convolutional Neural Networks Using Fine-tuned CP-Decomposition

Cited by in corpus (38)

• MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications
• Encoder-Decoder with Atrous Separable Convolution for Semantic Image Segmentation
• ENet: A Deep Neural Network Architecture for Real-Time Semantic Segmentation
• ShuffleNet: An Extremely Efficient Convolutional Neural Network for Mobile Devices
• Convolutional Networks for Fast, Energy-Efficient Neuromorphic Computing
• Xception: Deep Learning with Depthwise Separable Convolutions
• Wide Activation for Efficient and Accurate Image Super-Resolution
• MobileViT: Light-weight, General-purpose, and Mobile-friendly Vision Transformer
• Hardware Acceleration of Sparse and Irregular Tensor Computations of ML Models: A Survey and Insights
• Review: Deep Learning in Electron Microscopy
• ACNet: Strengthening the Kernel Skeletons for Powerful CNN via Asymmetric Convolution Blocks
• Interleaved Group Convolutions for Deep Neural Networks
• MiniSeg: An Extremely Minimum Network for Efficient COVID-19 Segmentation
• ExpandNets: Linear Over-parameterization to Train Compact Convolutional Networks
• PDNet: Semantic Segmentation integrated with a Primal-Dual Network for Document binarization
• ANTNets: Mobile Convolutional Neural Networks for Resource Efficient Image Classification
• Efficient Differentiable Neural Architecture Search with Meta Kernels
• OpenEI: An Open Framework for Edge Intelligence
• Shift-based Primitives for Efficient Convolutional Neural Networks
• XSepConv: Extremely Separated Convolution
• SmartExchange: Trading Higher-cost Memory Storage/Access for Lower-cost Computation
• State Classification with CNN
• Real-time Memory Efficient Large-pose Face Alignment via Deep Evolutionary Network
• Learning Student Networks via Feature Embedding
• Gamified AI Approch for Early Detection of Dementia
• Deep Learning-based Aerial Image Segmentation with Open Data for Disaster Impact Assessment
• Adaptive Selection of Deep Learning Models on Embedded Systems
• LPRNet: Lightweight Deep Network by Low-rank Pointwise Residual Convolution
• DelugeNets: Deep Networks with Efficient and Flexible Cross-layer Information Inflows
• WSNet: Compact and Efficient Networks Through Weight Sampling
• Butterfly-Net2: Simplified Butterfly-Net and Fourier Transform Initialization
• HBONet: Harmonious Bottleneck on Two Orthogonal Dimensions
• Robust Student Network Learning
• Penetrating the Fog: the Path to Efficient CNN Models
• A Survey on Green Deep Learning
• Faster Convolution Inference Through Using Pre-Calculated Lookup Tables
• Multi-objective Evolutionary Approach for Efficient Kernel Size and Shape for CNN
• Asymmetric CNN for image super-resolution