CoDeNet: Efficient Deployment of Input-Adaptive Object Detection on Embedded FPGAs

Zhen Dong, Dequan Wang, Qijing Huang, Yizhao Gao, Yaohui Cai, Tian Li, Bichen Wu, K. Keutzer, J. Wawrzynek
{"title":"CoDeNet: Efficient Deployment of Input-Adaptive Object Detection on Embedded FPGAs","authors":"Zhen Dong, Dequan Wang, Qijing Huang, Yizhao Gao, Yaohui Cai, Tian Li, Bichen Wu, K. Keutzer, J. Wawrzynek","doi":"10.1145/3431920.3439295","DOIUrl":null,"url":null,"abstract":"Deploying deep learning models on embedded systems for computer vision tasks has been challenging due to limited compute resources and strict energy budgets. The majority of existing work focuses on accelerating image classification, while other fundamental vision problems, such as object detection, have not been adequately addressed. Compared with image classification, detection problems are more sensitive to the spatial variance of objects, and therefore, require specialized convolutions to aggregate spatial information. To address this need, recent work introduces dynamic deformable convolution to augment regular convolutions. Regular convolutions process a fixed grid of pixels across all the spatial locations in an image, while dynamic deformable convolution may access arbitrary pixels in the image with the access pattern being input-dependent and varying with spatial location. These properties lead to inefficient memory accesses of inputs with existing hardware. In this work, we harness the flexibility of FPGAs to develop a novel object detection pipeline with deformable convolutions. We show the speed-accuracy tradeoffs for a set of algorithm modifications including irregular-access versus limited-range and fixed-shape on a flexible hardware accelerator. We evaluate these algorithmic changes with corresponding hardware optimizations and show a 1.36x and 9.76x speedup respectively for the full and depthwise deformable convolution on hardware with minor accuracy loss. We then co-design a network called CoDeNet with the modified deformable convolution for object detection and quantize the network to 4-bit weights and 8-bit activations. With our high-efficiency implementation, our solution reaches 26.9 frames per second with a tiny model size of 0.76 MB while achieving 61.7 AP50 on the standard object detection dataset, Pascal VOC. With our higher-accuracy implementation, our model gets to 67.1 AP50 on Pascal VOC with only 2.9 MB of parameters--20.9x smaller but 10% more accurate than Tiny-YOLO.","PeriodicalId":386071,"journal":{"name":"The 2021 ACM/SIGDA International Symposium on Field-Programmable Gate Arrays","volume":"21 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2021-02-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"37","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"The 2021 ACM/SIGDA International Symposium on Field-Programmable Gate Arrays","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1145/3431920.3439295","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 37

Abstract

Deploying deep learning models on embedded systems for computer vision tasks has been challenging due to limited compute resources and strict energy budgets. The majority of existing work focuses on accelerating image classification, while other fundamental vision problems, such as object detection, have not been adequately addressed. Compared with image classification, detection problems are more sensitive to the spatial variance of objects, and therefore, require specialized convolutions to aggregate spatial information. To address this need, recent work introduces dynamic deformable convolution to augment regular convolutions. Regular convolutions process a fixed grid of pixels across all the spatial locations in an image, while dynamic deformable convolution may access arbitrary pixels in the image with the access pattern being input-dependent and varying with spatial location. These properties lead to inefficient memory accesses of inputs with existing hardware. In this work, we harness the flexibility of FPGAs to develop a novel object detection pipeline with deformable convolutions. We show the speed-accuracy tradeoffs for a set of algorithm modifications including irregular-access versus limited-range and fixed-shape on a flexible hardware accelerator. We evaluate these algorithmic changes with corresponding hardware optimizations and show a 1.36x and 9.76x speedup respectively for the full and depthwise deformable convolution on hardware with minor accuracy loss. We then co-design a network called CoDeNet with the modified deformable convolution for object detection and quantize the network to 4-bit weights and 8-bit activations. With our high-efficiency implementation, our solution reaches 26.9 frames per second with a tiny model size of 0.76 MB while achieving 61.7 AP50 on the standard object detection dataset, Pascal VOC. With our higher-accuracy implementation, our model gets to 67.1 AP50 on Pascal VOC with only 2.9 MB of parameters--20.9x smaller but 10% more accurate than Tiny-YOLO.
嵌入式fpga上输入自适应目标检测的有效部署
由于有限的计算资源和严格的能源预算,在嵌入式系统上部署深度学习模型用于计算机视觉任务一直具有挑战性。现有的大部分工作都集中在加速图像分类上,而其他基本的视觉问题,如目标检测,还没有得到充分的解决。与图像分类相比,检测问题对物体的空间变化更为敏感,因此需要专门的卷积来聚合空间信息。为了满足这一需求,最近的工作引入了动态可变形卷积来增强正则卷积。规则卷积处理图像中所有空间位置上的固定像素网格,而动态可变形卷积可以访问图像中的任意像素,其访问模式依赖于输入并随空间位置而变化。这些属性导致使用现有硬件对输入进行低效的内存访问。在这项工作中,我们利用fpga的灵活性来开发一种具有可变形卷积的新型目标检测管道。我们展示了一组算法修改的速度-精度权衡,包括不规则访问与灵活硬件加速器上的有限范围和固定形状。我们用相应的硬件优化来评估这些算法的变化,结果显示,在精度损失较小的情况下,硬件上的完全和深度可变形卷积的速度分别提高了1.36倍和9.76倍。然后,我们共同设计了一个名为CoDeNet的网络,该网络具有改进的可变形卷积,用于对象检测,并将网络量化为4位权重和8位激活。通过我们的高效实现,我们的解决方案以0.76 MB的微小模型尺寸达到每秒26.9帧,同时在标准目标检测数据集Pascal VOC上实现61.7 AP50。通过我们更高精度的实现,我们的模型在Pascal VOC上达到67.1 AP50,只有2.9 MB的参数-比Tiny-YOLO小20.9倍,但精度提高10%。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
求助全文
约1分钟内获得全文 求助全文
来源期刊
自引率
0.00%
发文量
0
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
确定
请完成安全验证×
copy
已复制链接
快去分享给好友吧!
我知道了
右上角分享
点击右上角分享
0
联系我们:info@booksci.cn Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。 Copyright © 2023 布克学术 All rights reserved.
京ICP备2023020795号-1
ghs 京公网安备 11010802042870号
Book学术文献互助
Book学术文献互助群
群 号:481959085
Book学术官方微信