TBN: Convolutional Neural Network with Ternary Inputs and Binary Weights
| Autor: | Li Liu, Ling Shao, Fan Zhu, Diwen Wan, Heng Tao Shen, Fumin Shen, Jie Qin |
|---|---|
| Rok vydání: | 2018 |
| Předmět: |
Computer science
Binary number 02 engineering and technology Pascal (programming language) 010501 environmental sciences 01 natural sciences Convolutional neural network Object detection Matrix multiplication Residual neural network Task (computing) Computer engineering 0202 electrical engineering electronic engineering information engineering 020201 artificial intelligence & image processing computer 0105 earth and related environmental sciences computer.programming_language |
| Zdroj: | Computer Vision – ECCV 2018 ISBN: 9783030012151 ECCV (2) |
| DOI: | 10.1007/978-3-030-01216-8_20 |
| Popis: | Despite the remarkable success of Convolutional Neural Networks (CNNs) on generalized visual tasks, high computational and memory costs restrict their comprehensive applications on consumer electronics (e.g., portable or smart wearable devices). Recent advancements in binarized networks have demonstrated progress on reducing computational and memory costs, however, they suffer from significant performance degradation comparing to their full-precision counterparts. Thus, a highly-economical yet effective CNN that is authentically applicable to consumer electronics is at urgent need. In this work, we propose a Ternary-Binary Network (TBN), which provides an efficient approximation to standard CNNs. Based on an accelerated ternary-binary matrix multiplication, TBN replaces the arithmetical operations in standard CNNs with efficient XOR, AND and bitcount operations, and thus provides an optimal tradeoff between memory, efficiency and performance. TBN demonstrates its consistent effectiveness when applied to various CNN architectures (e.g., AlexNet and ResNet) on multiple datasets of different scales, and provides \(\sim \)32\(\times \) memory savings and \(40\times \) faster convolutional operations. Meanwhile, TBN can outperform XNOR-Network by up to 5.5% (top-1 accuracy) on the ImageNet classification task, and up to 4.4% (mAP score) on the PASCAL VOC object detection task. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|