HOTCAKE: Higher Order Tucker Articulated Kernels for Deeper CNN Compression

Abstract

The emerging edge computing has promoted immense interests in compacting a neural network without sacrificing much accuracy. In this regard, low-rank tensor decomposition constitutes a powerful tool to compress convolutional neural networks (CNNs) by decomposing the 4-way kernel tensor into multi-stage smaller ones. Building on top of Tucker-2 decomposition, we propose a generalized Higher Order Tucker Articulated Kernels (HOTCAKE) scheme comprising four steps: input channel decomposition, guided Tucker rank selection, higher order Tucker decomposition and fine-tuning. By subjecting each CONV layer to HOTCAKE, a highly compressed CNN model with graceful accuracy trade-off is obtained. Experiments show HOTCAKE can compress even pre-compressed models and produce state-of-the-art lightweight networks.

arXiv comment6 pages, 5 figures

References in corpus (8)

• Improving neural networks by preventing co-adaptation of feature detectors
• Binarized Neural Networks: Training Deep Neural Networks with Weights and Activations Constrained to +1 or -1
• Exploiting Linear Structure Within Convolutional Networks for Efficient Evaluation
• Tensor Decompositions for Signal Processing Applications From Two-way to Multiway Component Analysis
• Ternary Weight Networks
• Pruning Filters for Efficient ConvNets
• Speeding-up Convolutional Neural Networks Using Fine-tuned CP-Decomposition
• Towards Principled Design of Deep Convolutional Networks: Introducing SimpNet