LCNN: Lookup-Based Convolutional Neural Network

By Hessam Bagherinezhad, Mohammad Rastegari, Ali Farhadi. 2016

96 citations recorded by the source corpus.

Porting state of the art deep learning algorithms to resource constrained compute platforms (e.g. VR, AR, wearables) is extremely challenging. We propose a fast, compact, and accurate model for convolutional neural networks that enables efficient learning and inference. We introduce LCNN, a lookup-based convolutional neural network that encodes convolutions by few lookups to a dictionary that is trained to cover the space of weights in CNNs. Training LCNN involves jointly learning a dictionary and a small set of linear combinations. The size of the dictionary naturally traces a spectrum of trade-offs between efficiency and accuracy. Our experimental results on ImageNet challenge show that LCNN can offer 3.2x speedup while achieving 55.1% top-1 accuracy using AlexNet architecture. Our fastest LCNN offers 37.6x speed up over AlexNet while maintaining 44.3% top-1 accuracy. LCNN not only offers dramatic speed ups at inference, but it also enables efficient training. In this paper, we show the benefits of LCNN in few-shot learning and few-iteration learning, two crucial aspects of on-device training of deep learning models.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computer Vision and Pattern Recognition

Citations in the corpus

Recent advances in convolutional neural networks (2015, background): “[192] propose a lookup-based convolutional neural network (LCNN) that encodes convolutions by few lookups to a rich set of dictionary that trained to cover the space of weights in CNNs.”
AMC: AutoML for Model Compression and Acceleration on Mobile Devices (2018, background): “10, 41] and special convolution implementations [36, 48, 29, 3] can also speed up the neural networks.”
Channel Pruning for Accelerating Very Deep Neural Networks (2017, methodology): “Optimized implementation based methods [36, 48, 27, 4] accelerate convolution, with special convolution algorithms like FFT [48].”
Adaptive Neural Networks for Efficient Inference (2017, background): “Reduced precision networks (Gong et al., 2014; Cour-bariaux et al., 2015; Chen et al., 2015; Hubara et al., 2016a; Wu et al., 2016; Rastegari et al., 2016; Hubara et al., 2016b) have been extensively studied to reduce the memory foot-print of networks and their test-time cost.”
ADC: Automated Deep Compression and Acceleration with Reinforcement Learning (2018, unknown): “Quantization [63, 12, 48] and special conv implementation [42, 55, 33, 4] speed up CNN inherently.”
The History Began from AlexNet: A Comprehensive Survey on Deep Learning Approaches (2018, methodology): “Look up based CNN[76] is a fast, compact, and accurate model enabling efficient inference.”
ESPNet: Efficient Spatial Pyramid of Dilated Convolutions for Semantic Segmentation (2018, methodology): “Sparse CNN: To remove the redundancy in CNNs, sparse CNN methods, such as sparse decomposition [32], structural sparsity learning [33], and dictionary-based method [34], have been proposed.”
ShuffleNet: An Extremely Efficient Convolutional Neural Network for Mobile Devices (2017, methodology): “Without modifying the parameters, optimized convolution algorithms implemented by FFT [25, 35] and other methods [2] decrease time consumption in practice.”
Towards Evolutionary Compression (2018, background): “[2] decomposed original convolution filters as weighted combinations of basis filters and sparse coefficients.”
Exploring the Regularity of Sparse Structure in Convolutional Neural Networks (2017, methodology): “For CNN models, we only count the overall sparsity of convolutional layers since convolutional layers take up most of the computations in a typical CNN model[24].”
MiniRocket: A Very Fast (Almost) Deterministic Transform for Time Series Classification (2020, methodology): “ly simplied versions of shift-based methods [37], where conventional convolutional kernels are replaced by a combination of 1 1 convolutions and spatial shifts in the input, and lookup-based methods [1], where the convolution operation is performed via linear combinations of the precomputed convolut”