A new algorithm for extracting common kernels and convolutions to maximally eliminate the redundant operations among the convolutions in binary- and ternary-weight convolutional neural networks is presented. Precisely, we propose (1) a new algorithm of common kernel extraction to overcome the local and limited exploration of common kernel candidates by the existing method, and subsequently apply (2) a new concept of common convolution extraction to maximally eliminate the redundancy in the convolution operations. In addition, our algorithm is able to (3) tune in minimizing the number of resulting kernels for convolutions, thereby saving the total memory access latency for kernels. Experimental results on ternary-weight VGG-16 demonstrate that our convolution optimization algorithm is very effective, reducing the total number of operations for all convolutions by 25.8∼26.3%, thereby reducing the total number of execution cycles on hardware platform by 22.4% while using 2.7∼3.8% fewer kernels over that of the convolution utilizing the common kernels extracted by the state-of-the-art algorithm.

中文翻译：

---

二元和三元权重神经网络中的常见内核和卷积

提出了一种提取公共核和卷积的新算法，以最大限度地消除二元和三元权重卷积神经网络中卷积之间的冗余操作。确切地说，我们建议（1）一种新的公共核提取算法克服现有方法对常见内核候选者的局部和有限探索，并随后应用（2）普通卷积提取的新概念最大限度地消除卷积操作中的冗余。此外，我们的算法能够 (3)调整最小化生成内核的数量用于卷积，从而节省内核的总内存访问延迟。三元权重 VGG-16 的实验结果表明，我们的卷积优化算法非常有效，将所有卷积的操作总数减少了25.8～26.3%，从而在使用时将硬件平台上的执行周期总数减少了 22.4%2.7～3.8%使用最先进的算法提取的通用内核，比卷积内核更少。