arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2104.14124cs.CVcs.AR
收藏

Condensation-Net:具有跨通道池层和虚拟特征映射的内存高效网络架构

Condensation-Net: Memory-Efficient Network Architecture with Cross-Channel Pooling Layers and Virtual Feature Maps

Canon Information Technology (Beijing) Co., Ltd.(佳能信息技术(北京)有限公司);Canon Inc.(佳能公司)

Tse-Wei Chen, Motoki Yoshinaga, Hongxing Gao, Wei Tao, Dongchao Wen, Junjie Liu, Kinya Osa, Masami Kato

摘要:

“轻量级卷积神经网络“是嵌入式视觉领域的一个重要研究课题。为了在资源有限的硬件平台上实现图像识别任务,有必要减少存储器大小和计算成本。本文的贡献如下。首先,我们提出了一种算法来处理一个特定的网络架构(凝聚网),而不增加最大内存存储的特征映射。通过在将特征图存储到内存中之前计算跨通道池化的结果,虚拟特征图的架构节省了26.5%的内存带宽。其次,我们表明,跨通道池可以提高目标检测任务的准确性,如人脸检测,因为它增加了过滤器权重的数量。与Tiny-YOLOv 2相比,当误报率为0.1时,量化网络的精度提高了2.0%,全精度网络的精度提高了1.5%。最后,分析结果表明,所提出的硬件架构支持跨通道池的开销可以忽略不计。支持Condensation-Net的额外内存成本是总大小的0.2%,而额外的门数仅为总大小的1.0%。

英文摘要:

"Lightweight convolutional neural networks" is an important research topic in the field of embedded vision. To implement image recognition tasks on a resource-limited hardware platform, it is necessary to reduce the memory size and the computational cost. The contribution of this paper is stated as follows. First, we propose an algorithm to process a specific network architecture (Condensation-Net) without increasing the maximum memory storage for feature maps. The architecture for virtual feature maps saves 26.5% of memory bandwidth by calculating the results of cross-channel pooling before storing the feature map into the memory. Second, we show that cross-channel pooling can improve the accuracy of object detection tasks, such as face detection, because it increases the number of filter weights. Compared with Tiny-YOLOv2, the improvement of accuracy is 2.0% for quantized networks and 1.5% for full-precision networks when the false-positive rate is 0.1. Last but not the least, the analysis results show that the overhead to support the cross-channel pooling with the proposed hardware architecture is negligible small. The extra memory cost to support Condensation-Net is 0.2% of the total size, and the extra gate count is only 1.0% of the total size.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑